Influence of substrate structure on PGA-catalyzed acylations. Evaluation of different approaches for the enzymatic synthesis of cefonicid

Article abstract of DOI:10.1002/adsc.200404136

The influence of the substrate structure on the catalytic properties of penicillin G acylase (PGA) from Escherichia coli in kinetically controlled acylations has been studied. In particular, the affinity of different β-lactam nuclei towards the active site has been evaluated considering the ratio between the rate of synthesis (v_s) and the rate of hydrolysis of the acylating ester (v_hl). 7-Aminocephalosporanic acid (7-ACA) and 7-amino-3-(1-sulfomethyl-1,2,3,4-tetrazol-5-yl)thiomethyl-3-cephem-4-carboxylic acid (7-SACA) showed a good affinity for the active centre of PGA. The enzymatic acylation of these nuclei with R-methyl mandelate has been studied in order to evaluate different approaches for the enzymatic synthesis of cefonicid. The best results have been obtained in the acylation of 7-SACA. Cefonicid (8) was recovered from the reaction mixture as the disodium salt in 65% yield and about 95% of purity. Furthermore, through acylation of 7-ACA, a "one-pot" chemo-enzymatic synthesis was carried out starting from cephalosporin C using three enzymes in sequence: D-amino acid oxidase (DAO), glutaryl acylase (GA) and PGA. Cefonicid disodium salt was obtained in three steps, avoiding any intermediate purification, in 35% overall yield and about 94% purity. This approach presents several advantages compared with the classical chemical processes.

Full text of DOI:10.1002/adsc.200404136

FULL PAPERS
Influence of Substrate Structure on PGA-Catalyzed Acylations.
Evaluation of Different Approaches for the Enzymatic Synthesis
of Cefonicid
Marco Terreni,a,* Joseph Gapesie Tchamkam,a Umberto Sarnataro,b, d
b
c
c
´
´
´
Silvia Rocchietti, Roberto Fernandez-Lafuente, Jose M. Guisan
a
`
Pharmaceutical Biocatalysis Laboratories, Dipartimento di Chimica Farmaceutica, Via Taramelli 12, Universita degli
Studi, 27100 Pavia, Italy
Phone: (þ39)-382-987265, fax: (þ39)-382-987589, e-mail: markt@pbl.unipv.it
b
c
Innovate Biotechnology s.r.l., Parco Scientifico Tecnologico, Strada Savonesa 9, 15050 Rivalta Scrivia (AL), Italy
´
´
Departamento de Biocatalisis, Instituto de Catalisis-CSIC Campus UAM, Cantoblanco, 28049 Madrid, Spain
Present address: Merck Sharp & Dohme Italia, Via Emilia 21, 27100 Pavia, Italy
d
Received: May 13, 2004; Revised: October 29, 2004; Accepted: November 12, 2004
Abstract: The influence of the substrate structure on was recovered from the reaction mixture as the diso-
the catalytic properties of penicillin G acylase dium salt in 65% yield and about 95% of purity. Fur-
(PGA) from Escherichia coli in kinetically controlled thermore, through acylation of 7-ACA, a “one-pot”
acylations has been studied. In particular, the affinity chemo-enzymatic synthesis was carried out starting
of different b-lactam nuclei towards the active site from cephalosporin C using three enzymes in se-
has been evaluated considering the ratio between quence: d-amino acid oxidase (DAO), glutaryl acylase
the rate of synthesis (vs) and the rate of hydrolysis (GA) and PGA. Cefonicid disodium salt was ob-
of the acylating ester (vh1). 7-Aminocephalosporanic tained in three steps, avoiding any intermediate puri-
acid (7-ACA) and 7-amino-3-(1-sulfomethyl-1,2,3,4- fication, in 35% overall yield and about 94% purity.
tetrazol-5-yl)thiomethyl-3-cephem-4-carboxylic acid This approach presents several advantages compared
(7-SACA) showed a good affinity for the active centre with the classical chemical processes.
of PGA. The enzymatic acylation of these nuclei with
R-methyl mandelate has been studied in order to
evaluate different approaches for the enzymatic syn- Keywords: d-amino acid oxidase; biotransformations;
thesis of cefonicid. The best results have been ob- cefonicid; enzyme catalysis; glutaryl acylase; penicillin
tained in the acylation of 7-SACA. Cefonicid (8) G acylase
Introduction
such as penicillin G acylase (PGA) from Escherichia
coli has been considered by many authors. For example,
Cefazolin, cefamandole and cefonicid are cephalospor- the enzymatic N-acylations of 7-TACA with R-methyl
ins of clinical relevance.[1] These compounds are cur- mandelate (MAME, 1)[5] and of 7-ZACA with tetra-
rently synthesized[2–4] through the cleavage of cephalo- zol-1-yl-2-acetic acid methyl ester (TAME, 2)[6] cata-
sporin C to give 7-aminocephalosporanic acid (7- lyzed by this enzyme have been reported for the synthe-
ACA) that is then reacted with the appropriate S-nucle- ses of cefamandole and cefazolin, respectively. Howev-
ophile affording 7-amino-3-(1-methyl-1,2,3,4-tetrazol- er, this approach only solves the problems relevant to
5-yl)thiomethyl-3-cephem-4-carboxylic acid (7-TACA), the N-acylation of the 3’-functionalized nuclei and, be-
7-amino-3-(1-sulfomethyl-1,2,3,4-tetrazol-5-yl)thiome- sides, the yields are often not competitive with the chem-
thyl-3-cephem-4-carboxylic acid (7-SACA) or 7-amino- ical process. To the best of our knowledge, no reports
3-(5-methyl-1,3,4-thiadiazol-2-yl)thiomethyl-3-cephem- have been published so far describing the synthesis of ce-
4-carboxylic acid (7-ZACA). Finally, the chemical acy- fonicid by enzymatic N-acylation of 7-SACA.
lation of the C-7 amino group gives, respectively, cefa-
The catalytic properties of the PGA from E. coli
mandole, cefonicid or cefazolin (Scheme 1). However, strongly depend on changes induced by the adsorption
this multi-step process requires the use of toxic reagents of the acyl donor, as suggested by crystallographic anal-
and solvents, the purification of the intermediates and yses,[7] as well as by the different behaviour of phenyl-
very low reaction temperatures (ꢀ408C). To partially acetic acid analogues used in the acylation reactions.[8,9]
avoid these drawbacks, the use of enzymatic catalysts Besides, also the b-lactam nucleus used as substrate can
Adv. Synth. Catal. 2005, 347, 121–128
DOI: 10.1002/adsc.200404136
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
121

Marco Terreni et al.
FULL PAPERS
Scheme 1.
affect the catalytic properties of this enzyme, for exam- was synthesized by N-acylation of 7-ACA with
ple, 7-ACA has been reported to be a better substrate MAME (1).[12]
than 6-aminopenicillanic acid[10] or 7-ZACA.[11]
In this work, the properties of PGA from E. coli have
Considering the better affinity of 7-ACA for the PGA been investigated in the kinetically controlled N-acyla-
active site, as well as its higher water solubility compared tion of different cephalosporanic nuclei with MAME
with 7-ZACA (the cefazolin nucleus), a new approach (1) and TAME (2). The influence of the substrate struc-
has been proposed for the synthesis of cefazolin. 7- ture on the enzymatic properties has been evaluated. In
ACA was prepared by enzymatic cleavage of cephalo- particular, the enzymatic N-acylations of 7-SACA and
sporin C, catalyzed by d-amino acid oxidase (DAO) 7-ACAwith 1 have been studied to exploit different ap-
and glutaryl acylase (GA) followed by PGA-catalyzed proaches for the synthesis of cefonicid. In fact, the enzy-
N-acylation of this nucleus with TAME (2) to obtain ce- matic synthesis of this cephalosporin has been studied
fazolin.[11] This approach presents several advantages considering both the N-acylation directly performed
compared with the classical chemical synthesis. The sub- on the isolated pure 7-SACA and the synthesis per-
strate specificity of the enzymes allowed us to prepare formed via PGA-catalyzed N-acylation of 7-ACA ac-
cefazolin without any intermediate isolation or purifica- cording to the “multienzymatic approach” previously
tion step. All reactions were carried out in fully aqueous developed for cefazolin and cefamandole.[11,12]
medium and mild conditions. Similarly, cefamandole
122
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 121–128

Approaches for the Enzymatic Synthesis of Cefonicid
FULL PAPERS
Table 1. Evaluation of synthetase/esterase activities ratio (vs/
vh1): influence of the b-lactam and acyl donor structures.
Results and Discussion
Acyl donor
Nucleus
vs[a]
vs/vh1
Product
Influence of the Substrate Structures on the Catalytic
Properties of PGA
1
1
1
1
2
2
2
7-ADCA
7-ACA
7-TACA
7-SACA
7-ADCA
7-ACA
0.45
>20
5
6
7
8
9
16
19
15
0.2
7
8
8.2
3.4
11.7
12.5
7.3
In the kinetically controlled N-acylation, the yields de-
pend on the balance of three different reactions cata-
lyzed by the same enzyme:[5,13] the synthesis (s), the hy-
drolysis of the activated acyl donor (h1) and the hydrol-
ysis of the synthesized antibiotic (h2). To obtain high
yields, the values of the ratio between the rate of syn-
thesis and the rates of product (vs/vh2) and ester (vs/vh1)
hydrolyses, should be as high as possible. Particularly,
the vs/vh1 ratio is strictly related to the affinity of the en-
zyme active centre for the b-lactam nucleus and it de-
10
11
7-TACA
2.1
Reaction conditions: T¼48C; pH 6.5; [b-lactam nucleus]:
50 mM; [acyl donor]: 10 mM.
[a]
vs is expressed as mmol/min/g (of enzyme derivative)
fines the yield that could be achieved when the hydroly- nature of the substituent in the 3’-position exerts on
sis of the reaction product is negligible.[5] This value de- the substrateꢁs affinity towards the PGA active site.
creases during the reaction course as far as the nucleus is
Using esters 1 and 2 (10 mM) for the N-acylation of 7-
consumed. At the beginning, the concentration of the aminodeacetoxycephalosporanic acid (7-ADCA, R¼
acylation product is very low and its hydrolysis can be H), 7-ACA (R¼OAc), 7-TACA (R¼1-methyl-1,2,3,4-
discarded. Thus, from the initial vs/vh1 ratio, the percent- tetrazol-5-mercaptyl) and 7-SACA (R¼1-sulfomethyl-
age of activated acyl donor transformed into the acyla- 1,2,3,4-tetrazol-5-mercaptyl), compounds 5–11 have
tion product can be evaluated (synthesis %).
been synthesized according to Scheme 2. We evaluated
Different cephalosporanic nuclei have been studied the initial vs/vh1 ratio monitoring the acids (3 and 4)
(Scheme 2) in order to evaluate the influence that the and the products formation at the beginning of the reac-
tion, as reported in the Experimental Section. In the N-
acylation of 7-ADCA (50 mM), a very high vs/vh1 value
(>20) was obtained in spite of the very low synthetic ac-
tivity observed with this nucleus (Table 1); the hydroly-
sis of the acylating ester 1 was, in fact, negligible.
With 7-ACA, 7-TACA and 7-SACA, higher synthetic
activities but lower initial vs/vh1 values were observed in
comparison with 7-ADCA. For example, the vs/vh1 for 7-
ACA was 8.2 and 3.4 for 7-TACA, although with both
nuclei the synthetic activity was remarkable (15–
20 mmol minꢀ1/g of enzyme derivative). A similar trend
was observed in the N-acylations performed with ester 2
(Table 1). In particular, the initial vs/vh1 ratios observed
with the different substrates decreased from 12.5 for 7-
ADCA to 7.3 and 2.1 for 7-ACA and 7-TACA, respec-
tively.
In Figure 1 is shown the percentage of synthesis (con-
sidered as percentage of ester transformed into the acy-
lation product at the beginning of the reaction) calculat-
ed at different concentrations of nucleus, from the initial
vs/vh1 values[5] obtained in the N-acylation of 7-ADCA,
7-ACA and 7-TACA with ester 1 (10 mM). The results
obtained confirm that modifications of the 3’-substitu-
ent sensitively affect the affinity of the b-lactam nucleus
towards the enzyme active site. The best curve (Fig-
ure 1) was obtained with 7-ADCA (having the less hin-
dered substituent, R¼H) in the synthesis of compound
5 (complete saturation at 20 mM of nucleus concentra-
tion). Similarly, 7-ACA (R¼OAc) proved to be a better
nucleophile than the 3’-functionalized nucleus (7-
Scheme 2.
TACA).
Adv. Synth. Catal. 2005, 347, 121–128
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
123

Marco Terreni et al.
FULL PAPERS
Figure 1. Acylation of 7-ADCA, 7-ACA and 7-TACA at different concentrations (Scheme 2) with 1 (MAME, 10 mM).
When 7-SACA was tested in the N-acylation with es-
ter 1 (Table 1) to cefonicid (8), the initial vs/vh1 was
11.7, much higher than that resulting with 7-ACA or 7-
TACA. The relevant difference of the values observed
with 7-TACA and 7-SACA is surprising as these 3’-nu-
clei differ only in the sulfonic group on the tetrazole
ring. Thus, these results seem to support that the changes
in the affinity observed cannot be ascribed only to the
steric hindrance of the substituent in the 3-position.
This conclusion appears to be confirmed by the crystal-
lographic analysis performed on the complex between
PGA and penicillin G sulfoxide (PGSO),[14] a non-hy-
drolyzable substrate analogue.[15] The structure of this
complex shows that PGSO is adsorbed in the active
site of the enzyme through the b-lactam ring and the car-
boxylic group of the thiazole ring. The two methyl
groups of the thiazole are instead placed on the open
side of the active site and seem to have a poor interaction
with the protein structure (Figure 2).
The position of these two groups in the penicillanic
structure could be considered quite indicative of the spa-
tial disposition of the substituent in the 3 position of sim-
Figure 2. Structure of the complex between PGA and penicil-
ple cephalosporanic substrates such as 7-ADCA. How- lin G sulfoxide (PGSO) obtained using the PDB file 1gm9.[14]
ever, considering more complex substrates like the 3’-
functionalized cephalosporanic nuclei, further studies complete saturation at the active centre (vs/vh1 >20) was
should be performed in order to understand the chemi- achieved at 100 mM concentration of the nucleus.
cal-physical parameters crucial in determining the affin-
ity of these substrates for the PGA active site.
In contrast, the acylating agent (1 or 2) seems not to Study of N-Acylations for the Enzymatic Synthesis of
influence the affinity of the b-lactam nucleus for the Cefonicid (8)
PGA active centre. Comparing the curves obtained in
the N-acylation of 7-ACAusing esters 1 and 2 (Figure 3), Several reactions have been studied using an excess of
the results were in fact very similar and, in both cases, a ester in an attempt to obtain the quantitative transfor-
124
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 121–128

Approaches for the Enzymatic Synthesis of Cefonicid
FULL PAPERS
Table 2. PGA-catalyzed acylation of different b-lactam nuclei.
Nucleus
Ester
Product
pH
MeOH [% v/v]
Conversion [%]
7-ACA
7-ACA
1
2
1
1
1
1
1
1
6
10
8
6
8
8
8
8
6.5
6.5
6.5
6.5
6.5
6.0
5.5
5.5
–
–
–
20
20
30
30
30
71
95
75
80
84
86
88
92
7-SACA
7-ACA
7-SACA
7-SACA
7-SACA
7-SACA[a]
Reaction conditions: T¼48C; [b-lactam nucleus]: 50 mM; [1] or [2]: 150 mM.
[a]
[7-SACA]: 100 mM; [1]: 300 mM
er than that obtained with 7-ACA, according to the bet-
ter affinity of this nucleus toward the active site of PGA.
The N-acylations of 7-ACA and 7-SACA have been
further studied in order to evaluate different approaches
for the enzymatic synthesis of cefonicid (8). This com-
pound can be directly obtained by N-acylation of 7-
SACA (Scheme 2) or, alternatively, using 6 obtained
from 7-ACA (Scheme 3) as precursor. Different condi-
tions have been considered to optimize the final yield
and the use of PGA covalently immobilized on agarose
beads by multipoint attachment ensured the complete
stability of the catalyst in a wide range of conditions, in
agreement with the results previously reported.[16]
In both cases, the conversion degree improved when
methanol was used as co-solvent, according to the effect
exerted by this solvent on the catalytic properties of
Figure 3. Acylation of 7-ACA at different concentrations PGA from E. coli.[10,17] Thus, in the presence of methanol
with 1 (MAME) and 2 (TAME).
(20% v/v), the conversion increased from 71% to 79% in
the N-acylation of 7-ACA, according to the vs/vh1 ratio
(from 5.5 to 7.5). Similarly, in the case of 7-SACA, yields
mation of the different nuclei into the corresponding improved from 75% to 84% (Table 2).
acylated compounds. N-Acylation of 7-ACA (50 mM)
The high water solubility of 7-SACA allowed us to test
with esters 1 and 2 (150 mM) afforded different results more “drastic” reaction conditions than those used with
(Table 2). In the synthesis of 6, yields were not quantita- other nuclei in an attempt to optimize the enzymatic N-
tive because of the high hydrolytic activity shown by the acylation. In this case, on decreasing the pH from 6.5 to
enzyme towards this compound. In fact, when the max- 5.5 and increasing the methanol concentration up to
imum conversion was achieved (about 70% in 2 hours), 30% (Table 2), the conversion was 88%. On increasing
the concentration of 6 rapidly decreased (results not also the concentration of 7-SACA (from 50 mM to
shown). On the contrary, the acylation of 7-ACA with 100 mM) the conversion degree further improved up
2 gave almost a quantitative conversion (about 95% in to 92%. In a preparative experiment, cefonicid (8) was
2 hours) as a consequence of the very low hydrolytic ac- recovered from the reaction mixture as its disodium
tivity shown by PGA towards the product 9 (the concen- salt in 65% yield and about 95% of purity (see Experi-
tration of this compound being constant after the maxi- mental Section).
mum yield was achieved).
According to Scheme 3, the “one-pot” synthesis of cefo-
Thus, in the N-acylation with 2, high yields can be ach- nicid (8) was instead performed starting from cephalospor-
ieved, as indicated by the vs/vh1 obtained in the condi- in C by the simultaneous use ofimmobilizedDAO and GA
tions used for this synthesis (vs/vh1 ¼5.0), while in the to obtain 7-ACA (96% yield). The enzymatic acylation of
acylation with 1, the high hydrolytic activity shown by this nucleus, catalyzed by PGA, was then performed di-
PGA towards the product makes it very difficult to rectly on the crude solution obtained after filtration of
reach yields as high as those indicated by the vs/vh1 value DAO and GA using methanol (20%) and a 3:1 molar ex-
(vs/vh1 ¼5.5). For the same reason, also in the reaction of cess of 1 (150 mM); this reaction afforded 6 in 80% yield.
7-SACA (Table 2) to synthesize 8 (cefonicid), the con- Afterfiltrationofthe immobilizedPGA, thefinal displace-
version was not quantitative (76%) although being high- ment of the 3-acetoxy group with 1-sulfomethyl 5-mercap-
Adv. Synth. Catal. 2005, 347, 121–128
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
125

Marco Terreni et al.
FULL PAPERS
The second strategy was a “one-pot” synthesis starting
from cephalosporin C performed using three different
enzymes (DAO, GA and PGA). This synthesis has
been accomplished through a PGA-catalyzed acylation
directly performed on the crude 7-ACA obtained by en-
zymatic cleavageof cephalosporin C. In spite ofthe yield
obtained, this process seems very advantageous if com-
pared with the synthesis performed by chemical or enzy-
matic N-acylation of the 3’-functionalized nucleus 7-
SACA. In particular, all reactions were performed in
aqueous medium, avoiding any intermediate purifica-
tion and with a 35% overall yield. With the aim to im-
prove the results of the enzymatic acylation step, differ-
ent approaches can be taken into account:
a) Selection of acylases with better catalytic proper-
ties;
b) Improvement of the catalytic properties of the acy-
lase selected by design of suitable immobilization
strategies;
Scheme 3.
c) Improvement of the performances of the processes
by reaction medium design.
totetrazole (SMT) was directly performed by heating the
aqueous solution of 6 at 658C (50% of conversion). In a
preparative experiment, the final product was isolated These studies are in progress and the results will be pre-
from the reaction mixture as its disodium salt (as reported sented in forthcoming publications.
in the Experimental Section) in 35% overall yield and
about 93–94% purity. The main impurity in the isolated
product was the residual SMT (about 2%).
Experimental Section
Conclusion
General
The results reported in the present work indicate that the
nature of the substituent at the C-3’ position of the ceph-
alosporanic nucleus may strongly influence the perform-
ance of PGA as catalyst in kinetically controlled acyla-
tions. The changes in the affinity observed for the PGA
Commercially available reagent grade solvents were used with-
out purification. Immobilized DAO and GA and the crude ex-
tract of PGA from E. coli were from Recordati S.p.a. (Opera,
Milano, Italy). Cross-linked 6% agarose beads (Sepharose 6B-
CL) were from Pharmacia Biotech AB (Uppsala, Sweden);
active site towards the different cephalosporanic nuclei MAME(1), 7-ACA and 7-ADCAwerepurchasedfromAldrich
(Milwaukee, WI, USA), whereas the other b-lactam nuclei and
1-H-tetrazol-1-ylacetic acid (TA) were from Farmabios S.r.l
(Gropello Cairoli, Pavia, Italy). The immobilization of acylase
on agarose beads by multiple-point attachment was performed
following the general procedure previously reported.[16]
The pH of the solutions during the enzymatic hydrolyses and
syntheses reactions were kept constant by automatic titration
using a 718 Stat Tritino from Metrohm (Herisau, Switzerland).
HPLC analyses were run on a Merck-Hitachi L-7100 equipped
with UV detector L-7400. The column was a LiChroCART
seem to be not related to the steric hindrance of the sub-
stituent at 3’ but they might be probably ascribed to elec-
troniceffectsand to thehydrophobicityofthemoiety. For
not functionalized substrates, such as 7-ADCA, this con-
clusion can be supported by a qualitative comparison
with the structure of the PGA-PGSO complex obtained
from crystallographic data.[14] For more complex sub-
strates like the 3’-functionalized cephalosporanic nuclei
considered in this work, further studies should be per-
formed in order to understand the chemical-physical 250-4 RP select-B (Merck, Darmstadt, Germany) kept at
258C using an L-7300 column oven.
parameters that can play a pivotal role in determining
their affinity for the PGA active site. In this context, fur-
therstudiesareinprogress, includingthe dockingofthese
nuclei in the PGA active site.
Considering the good affinity of 7-SACA and 7-ACA
towards the PGA active site, two approaches have been
proposed for the enzymatic synthesis of cefonicid. The
first approach was the PGA catalyzed acylation of the
3’-functionalized nucleus that gave the product in good
yield and purity.
Mass spectra were recorded with a Finnigan (San Jose, CA,
USA) LCQ ion trap mass spectrometer. HPLC elution was in-
troduced into the electrospray source at 300 mL·minꢀ1. The
ESI source was operated at 4.5 kV, the capillary temperature
was set at 2508C and its voltage at 10 eV; the experiment was
performed in the negative ion mode. MS/MS: the ion of interest
was isolated in the ion trap and collisionally activated with 25%
ejection amplitude at standard He pressure. 1H NMR spectra
were recorded with a Bruker 400 MHz AMX spectrometer
(chemical shifts are reported in ppm).
126
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 121–128

Approaches for the Enzymatic Synthesis of Cefonicid
FULL PAPERS
The acylation products 5–7 and 9–11 were identified by
HPLC-mass spectrometry considering their molecular mass:
5¼346.0; 6¼407.0; 7¼460.1; 9¼323.1; 10¼381.1; 11¼437.0.
Cefonicid (8) was identified by HPLC analysis by comparison
with a sample of “Cefonicid bisodic salt – USP Working stand-
ard”.
Synthesis of (1H-Tetrazol-1-yl)-2-acetic Acid Methyl
Ester (TAME, 2)
Tetrazol-1-ylacetic acid (TA, 4, 1.5 g) was refluxed in dry meth-
anol (50 mL) containing 96% H2SO4 (150 mL). After 6 hours,
the solvent was evaporated and the oil obtained was dissolved
in ethyl acetate (50 mL). The resulting solution was washed
(2ꢁ20 mL) with NaHCO3 solution. The organic phases were
dried over anhydrous Na2SO4, filtered and the solvent was re-
moved under vacuum affording pure 2; yield: 1.14 g (69%); mp
50–528C; 1H NMR (DMSO-d6): d¼3.81 (s, 3H), 5.66 (s, 2H),
9.48 (s, 1H); HPLC analysis (210 nm): 2.5% acetonitrile in
10 mM phosphate buffer, pH 3.2, flow 1.5 mL/min: TA (4)
Rt¼1.64 min; TAME (2) Rt¼4.47 min (purity>98%).
Study of the Enzymatic N-Acylations
The study of the acylation of 7-ACA and 7-SACA (50 mM) has
been performed under different reaction conditions, using
150 mM of ester 1. The nucleus was first dissolved in 25 mM
phosphate buffer (20 mL) and then 1 (0.498 g) was added to
the resultant solution. When the reaction was performed in
presence of methanol, the ester was added as a solution in
the amount of methanol necessary for the reaction. After the
complete dissolution of the substrates, the mixture was cooled
(48C) and the pH adjusted at the desired value. A suitable
amount of immobilized PGA from E. coli was then added
(0.5–1 g) under magnetic stirring (200 rpm) keeping the pH
constant by automatic titration. The reactions were monitored
by HPLC analysis at 274 nm using the analytical conditions re-
ported above for compounds 6 and 8.
Determination of the Esterasic and Amidasic
Activities (ve/va)
The esterasic and amidasic activities of PGA have been evalu-
ated by measuring the initial hydrolysis rate of esters 1 or 2 and
of the corresponding cephalosporin. As a general procedure,
the enzyme derivative (20 UI) was added to the substrate sol-
ution (20 mL, 10 mM) under magnetic stirring (200 rpm). Dur-
ing the hydrolysis the pH was kept constant by automatic titra-
tion and the activities were evaluated from the sodium hydrox-
ide (100 mM) consumption (mmol of substrate transformed/
min/g of enzyme derivative).
Enzymatic Synthesis of Cefonicid (8) by Enzymatic
N-Acylation of 7-SACAþ
Determination of vs and vh1 in the N-Acylation of
b-Lactam Nuclei and the Saturation Curves
In a general procedure, 7-SACA (4.08 g) was dissolved in
25 mM phosphate buffer (160 mL) at pH 7. Ester 1 (4.98 g),
dissolved in methanol (60 mL, 30% of the total volume) was
then added and the pH was adjusted to 5.5. Immobilized
PGA (10 g) was added to the reaction mixture under mechan-
ical stirring. The reaction was carried out as reportedabove and
monitored by HPLC analysis at 274 nm using the analytical
conditions reported above for 8. When the maximum conver-
sion of 7-SACA into cefonicid (8) had been achieved, the enzy-
matic catalyst was recovered by filtration. The crude solution
was concentrated under vacuum and acidified to pH 2.5 using
3 N HCl, maintaining the mechanical stirring and the temper-
ature at 48C. The acidic solution was extracted with ethyl ace-
tate (3ꢁ150 mL). The aqueous solution was further acidified
(pH 1.5) and saturated with NaCl (80 g). The mixture was
then extracted with THF (3ꢁ80 mL). The organic extracts
were collected and washed twice with brine (50 mL). The or-
ganic phases were evaporated under reduced pressure keeping
the temperature below 328C. The resultant oil was dissolved in
ethanol (50 mL) and further concentrated under vacuum. Fi-
nally, the raw product was dissolved in a mixture of acetone/
ethanol(2:1, 90 mL) and added dropwise to a solution contain-
ing sodium 2-ethylhexanoate (2.5 g in 45 mL of acetone/etha-
nol 2:1) at 48C. The white precipitate obtained was filtered
and dried under vacuum. Cefonicid bisodium salt was obtained
in 65% yield (3.8 g). The product was analyzed by HPLC at 220
and 254 nm using the analytical conditions above reported for
compound 8. Identification was performed by comparing the
isolated product with “Cefonicid bisodic salt – USP Working
standard”. The purity of the isolated 8 as evaluated by HPLC
analysis was 94% at 220 nm and 96% at 254 nm.
The vs/vh1 ratios in the PGA-catalyzed acylations of the differ-
ent b-lactam nuclei have been evaluated by reaction with ester
1 or 2 at 10 mM concentration. The reactions were performed
at the desired concentration of the b-lactam nucleus (48C,
pH 6.5). Following a general procedure, the ester was dissolved
into a solution of the b-lactam nucleus (20 mL in 10 mM phos-
phate buffer) at the desired concentration. The enzyme deriv-
ative of PGA (20 UI) was then added to the solution at 48C un-
der magnetic stirring (200 rpm). During the reaction, the pH
was kept constant by automatic titration. The vs/vh1 ratio was
evaluated by measuring the initial rates of the syntheses (vs)
and ester hydrolyses (vh1) before 20% of the initial amount of
ester had been consumed. The acids and products formation
was monitored by HPLC analysis at 210 nm. The curves ob-
tained with the different b-lactam nuclei (Figures 1 and 3)
were plotted by performing the acylation reactions using a con-
stant concentration (10 mM) of ester, while increasing the con-
centration of the nucleus and recording the % of synthesis
(100ꢁvs/vs/vh1) versus the nucleus concentration. In the acyla-
tion with MAME (1), the analytical conditions were 20% ace-
tonitrile in 10 mM phosphate buffer, pH 3.2, flow 1.0 mL/min:
MA (3), Rt¼3.45 min; 5, Rt¼8.2 min; 6, Rt¼9.40 min; 7, Rt¼
19 min; 1, Rt¼12.6 min. For compound 8, obtained by acyla-
tion of 7-SACA, the HPLC analysis was performed with
12.5% acetonitrile (8, Rt¼6.7 min). In the acylation with
TAME (2), the conditions used were 2.5% acetonitrile in
10 mM phosphate buffer, pH 3.2, flow 1.5 mL/min: TA (4)
Rt¼1.64 min; 2, Rt¼4.47 min; 9, Rt¼8.0 min; 10, Rt¼
10.0 min; 11, Rt¼13.0 min.
Adv. Synth. Catal. 2005, 347, 121–128
asc.wiley-vch.de
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
127

Marco Terreni et al.
FULL PAPERS
ics, Academic Press, New York, 1982, pp. 371–437;
b) M. F. Rein, F. B. Westervelt, M. A. Sande, Antimicrob.
Agents Chemother. 1973, 4, 366–371; c) M. Nishida, T.
Murakawa, T. Kamimura, N. Okada, H. Sakamoto, S. Fu-
kada, S. Nakamoto, Y. H. Yokota, K. Miki, Antimicrob.
Agents Chemother. 1976, 10, 1–13; d) T. Noto, T. Neha-
shi, H. Endo, M. Saito, S. Matsubara, Y. Harada, S. Suzu-
ki, H. Ogawa, K. Koyama, J. Antibiot. 1976, 29, 1058–
1070.
Multienzymatic Synthesis of Cefonicid (8)
To a solution of cephalosporin C (4.15 g) in 25 mM phosphate
buffer pH 8 (160 mL) at 258C the enzyme derivatives of DAO
(300 UI) and GA (500 UI) were added under magnetic stirring
(200 rpm). During the reaction, a continuous flow of O2 was
maintained and the pH was kept constant by automatic titra-
tion. The reaction was monitored by HPLC (3% acetonitrile,
20 mM ammonium acetate buffer, pH 3.2, flow 1.5 mL/min; l
274 nm). After 3 hours, the complete deacylation of cephalo-
sporin C into 7-ACA (Rt¼3.24 min) was achieved and the re-
action mixture was filtered. The crude solution of 7-ACA was
used for the next reaction by addition of 1 (4.98 g) dissolved
in methanol (40 mL). The mixture was cooled to 48C and the
pH adjusted at 6.5. Immobilized PGA (10 g) was then added
under magnetic stirring (200 rpm) keeping the pH constant
by automatic titration. The reactions were monitored by
HPLC analysis at 274 nm using the analytical conditions re-
ported above for 6. When the maximum yield had been ach-
ieved, the reaction mixture was filtered. The solution of the
crude product 6, obtained after filtration of the immobilized
PGA, was used without purification for 3’-functionalization
to obtain cefonicid (8). In this reaction, SMT (3.92 g) was dis-
solved in the crude solution of 6 obtained after enzymatic acy-
lation. The mixture was heated at 658C at pH 6.5 for 4 hours af-
fording 8 in 50% yield as evaluated by HPLC analysis per-
formed as reported above at 274 nm and using an external
standard of “Cefonicid bisodic salt – USP Working standard”.
From the aqueous solution obtained after the last chemical
step, cefonicid was isolated as its disodium salt prepared with
sodium 2-ethylhexanoate in organic medium following the
procedure reported above but performing the extraction with
a larger volume of THF (6ꢁ80 mL). After concentration of
the THF extracts, the sodium salt was prepared dissolving the
oil obtained in 90 mL of a mixture of acetone/ethanol (2:1)
and dropping this solution into a solution of 2.5 g of sodium
2-ethylhexanoate in 45 mL of acetone/ethanol (2:1) cooled
at 48C. The white precipitate obtained was filtered and dried
under vacuum. Cefonicid bisodium salt was obtained in 35%
yield (2.0 g). The product was analyzed by HPLC as reported
above for the enzymatic synthesis. The purity of the isolated ce-
fonicid (8) as evaluated by HPLC analysis was 92% at 220 nm
and 94% at 254 nm.
[2] a) L. Sogli, D. Terrassan, G. Ribaldone, US Patent
5,387,679, 1995; b) W. Dꢂrckheimer, J. Blumbach, R.
Lattrel, K. H. Scheunemann, Angew. Chem. Int. Ed.
Engl. 1985, 24, 180–202.
[3] T. J. Polansky, US Patent 4,576,937, 1986.
[4] M. Brenner, V. Renganathan, US Patent WO 92/17600,
1992.
´
[5] O. Hernandez-Justiz, M. Terreni, G. Pagani, J. L. Garcꢃa-
´
´
´
Lopez, J. M. Guisan, R. Fernandez-Lafuente, Enzyme
Microb. Technol. 1999, 25, 336–343.
[6] a) J. I. Won, C. G. Kim, J. H. Kim, J. H. Lee, Y. J. Jeon,
Appl. Biochem. Biotechnol. 1998, 69, 1–9; b) M. Kostadi-
nov, A. Nikolov, N. Tsoneva, N. Petkov, Appl. Biochem.
Biotechnol. 1992, 33, 177–182.
[7] S. H. Done, J. A. Branningam, P. C. E. Moody, R. E. Hu-
bard, J. Mol Biol. 1998, 284, 463–475.
[8] R. Fernandez-Lafuente, C. M. Rossel, J. M. Guisan, En-
zyme Microb. Technol. 1998, 22, 583–587.
[9] N. Stambolieva, Z. Mincheva, B. Galunsky, Biocat. and
Biotrans. 1998, 16, 225–232.
´
´
[10] R. Fernandez-Lafuente, C. M. Rosell, J. M. Guisan, En-
zyme Microb. Technol. 1998, 23, 305–310.
´
`
[11] O. Hernandez-Justiz, R. Fernandez-Lafuente, J. M. Gui-
`
san, P. Negri, G. Pagani, M. Pregnolato, M. Terreni, J.
Org. Chem. 1997, 62, 9099–9106.
[12] M. Terreni, G. Pagani, D. Ubiali, R. Fernandez-Lafuente,
C. Mateo, J. M Guisan, Biorg. Med. Chem. Letter 2001,
11, 2429–2432.
[13] a) V. Kasche, Enzyme Microb. Technol. 1986, 8, 4–16;
b) V. Kasche, U. Haufler, L. Riechman, Meth. Enzymol.
1987, 136, 280–292.
[14] a) C. E. McVey,. M. A. Walsh, G. G. Dodson, K. S. Wil-
son, J. A. Brannigan, J. Mol. Biol. 2001, 313, 139–150;
b) the freeware WebLab Viewer Lite (http://www.msi.
com) was used to depict the structure of the complex.
[15] A. Plaskie, E. Roets, H. Vanderhaeghe, J. Antibiotics
1978, 31, 783–788.
Acknowledgements
We thank Farmabios S.r.l. (Gropello Cairoli, PV, Italy) for fi-
nancial support and Recordati S.p.a. (Opera, MI, Italy) for the
enzymes. We thank Dr. A. Rampino for his help in the early
stages of the project and Dr. D. Ubiali for helpful discussion.
This work has been partially funded by the Spanish CICYT
(projects BIO2001-2259 and PPQ 2002-01231)
´
´
[16] a) J. M. Guisan, R. Fernandez-Lafuente, V. Rodriguez,
A. Bastida, R. M. Blanco, G. Alvaro, in: Stability and Sta-
bilization of Enzymes, Elsevier Science Publishers, The
´
Netherlands, 1993, pp. 55–62; b) G. Alvaro, R. Fernan-
´
dez-Lafuente, R. M. Blanco, J. M. Guisan, Appl. Bio-
chem. Biotechnol. 1990, 26, 181–195; c) J. M. Guisan,
Enzyme Microb. Technol. 1988, 10, 375–382.
[17] a) M. G. Kim, S. B. Lee, J. Mol. Cat. B: Enzym. 1996, 1,
71–80; b) M .G. Kim, S. B. Lee, J. Mol. Cat. B: Enzym.
1996, 1, 181–190; c) M. G. Kim, S. B. Lee, J. Mol. Cat.
B: Enzym. 1996, 1, 201–211.
References and Notes
[1] a) J. A. Webber, W. J. Wheeler, Antimicrobial and phar-
macokinetic properties of newer penicillins and cephalo-
sporins. In: Chemistry and Biology of b-Lactam Antibiot-
128
ꢀ 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2005, 347, 121–128