Synthesis and in vitro evaluation of macromolecular antitumour derivatives based on phenylenediamine mustard

Article abstract of DOI:10.1016/j.ejps.2004.09.006

Poly-[N-(2-hydroxyethyl)-l-glutamine] (PHEG) and poly(ethylene glycol) (PEG)-grafted PHEG conjugates of N,N-di(2-chloroethyl)-4-phenylenediamine mustard (PDM) were synthetised. A collagenase-sensitive oligopeptide spacer was selected to link the cytotoxic agent PDM onto the polymeric carrier. First, the oligopeptide-drug conjugate, l-pro-l-leu-gly-l-pro-gly-PDM, was prepared. In a second step, the low molecular weight PDM derivative and PEG-NH₂ were coupled to a N,N-disuccinimidylcarbonate activated PHEG. Dynamic laser light scattering measurements indicated the formation of aggregates. The presence of human serum albumin had no significant effect on the diameter of the conjugates. The hydrolytic stability of the conjugates was investigated in buffer solutions. The conjugates showed an improved stability compared to the parent nitrogen mustard. The enzymatic degradation studies of the polymeric conjugates were performed in the presence of collagenase type IV (Clostridiopeptidase A; EC 3.4.24.3), cathepsin B (EC 3.4.22.1), cathepsin D (EC 3.4.23.5) and tritosomes. Only the bacterial collagenase type IV was able to cleave the spacer releasing free PDM and its peptidyl derivative, gly-l-pro-gly-PDM. The in vitro cytotoxicity of the conjugates was evaluated against HT1080 fibrosarcoma cells and MDA adenocarcinoma cells. All conjugates showed low toxicity towards these cell lines.

Full text of DOI:10.1016/j.ejps.2004.09.006

European Journal of Pharmaceutical Sciences 24 (2005) 159–168
Synthesis and in vitro evaluation of macromolecular antitumour
derivatives based on phenylenediamine mustard
Katleen De Winnea, Leonard W. Seymourb, Etienne H. Schachta,∗
a
Polymer Materials Research Group, Ghent University, Krijgslaan 281 S4-bis, 9000 Ghent, Belgium
b
Department of Clinical Pharmacology, Radcliffe Infirmary, University of Oxford, Woodstock Road, Oxford OX2 6HE, UK
Received 3 February 2004; received in revised form 3 September 2004; accepted 15 September 2004
Available online 25 December 2004
Abstract
Poly-[N-(2-hydroxyethyl)-l-glutamine] (PHEG) and poly(ethylene glycol) (PEG)-grafted PHEG conjugates of N,N-di(2-chloroethyl)-4-
phenylenediamine mustard (PDM) were synthetised. A collagenase-sensitive oligopeptide spacer was selected to link the cytotoxic agent
PDM onto the polymeric carrier. First, the oligopeptide-drug conjugate, l-pro-l-leu-gly-l-pro-gly-PDM, was prepared. In a second step, the
low molecular weight PDM derivative and PEG-NH2 were coupled to a N,N-disuccinimidylcarbonate activated PHEG. Dynamic laser light
scattering measurements indicated the formation of aggregates. The presence of human serum albumin had no significant effect on the diameter
of the conjugates. The hydrolytic stability of the conjugates was investigated in buffer solutions. The conjugates showed an improved stability
compared to the parent nitrogen mustard. The enzymatic degradation studies of the polymeric conjugates were performed in the presence
of collagenase type IV (Clostridiopeptidase A; EC 3.4.24.3), cathepsin B (EC 3.4.22.1), cathepsin D (EC 3.4.23.5) and tritosomes. Only the
bacterial collagenase type IV was able to cleave the spacer releasing free PDM and its peptidyl derivative, gly-l-pro-gly-PDM. The in vitro
cytotoxicity of the conjugates was evaluated against HT1080 fibrosarcoma cells and MDA adenocarcinoma cells. All conjugates showed low
toxicity towards these cell lines.
© 2004 Published by Elsevier B.V.
Keywords: Polymeric conjugate; Phenylenediamine mustard; Collagenase; Poly(hydroxyalkylglutamine)
1. Introduction
The main objective of covalent linking a conventional an-
titumour agent to a macromolecular carrier is to improve
Abbreviations: PDM, N,N-di(2-chloroethyl)-4-phenylenediamine mus-
the therapeutic index of the parent drug. The concept of
tard; PHEG, poly-[N-(2-hydroxyethyl)-l-glutamine]; PEG, poly(ethylene
water-soluble macromolecular prodrugs was first proposed
glycol); Z-, benzyloxycarbonyl-; MMC, mitomycin C; EDTA, ethylenedi-
by Ringsdorf (1975). Such conjugates are designed to circu-
late in the bloodstream for an extended period of time fol-
lowed by accumulation within the tumour tissue (Maeda and
Matsumura, 1989).
A site-specific drug release can be controlled by select-
ing a proper drug–polymer linkage. During circulation in the
bloodstream the linkage should be stable, but in or near the
target cells, e.g. cancer cells, the spacer should be degraded,
resultinginthereleaseoftheantitumouragent. Consequently,
conjugates having oligopeptide spacers, which are substrates
for lysosomal and tumour-associated enzymes have been
aminetetraacetic acid; NMR, nuclear magnetic resonance; s, singlet; bs,
broad singlet; d, doublet; dd, doublet of doublet; t, triplet; m, multi-
plet; DCC, dicyclohexylcarbodiimide; DCU, diclohexylureum; NHS, N-
hydroxysuccinimide; TLC, thin layer chromatography; Rf, ratio to front;
EtAc, ethylacetate; AcOH, acetic acid; GABA, ␥-aminobutyric acid; DSC,
N,N-disuccinimidylcarbonate; ISA, ionic strength adjustor; PBS, phosphate
buffered saline; DLS, dynamic light scattering; HPLC, high performance
liquid chromatography; DME, Dulbecco’s Modified Eagle medium; MTT,
3-(4,5-dimethyl)thiazol-2,5-diphenyl tetrazolium bromide; DMSO, N,N-
dimethylsulfoxide; GPC, gel permeation chromatography; Mw, weight av-
era∗ge molecular weight; MMP, matrix metalloprotease
Corresponding author. Tel.: +32 9 264 44 97; fax: +32 9 264 49 72.
E-mail address: etienne.schacht@ugent.be (E.H. Schacht).
0928-0987/$ – see front matter © 2004 Published by Elsevier B.V.
doi:10.1016/j.ejps.2004.09.006

160
K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
Another possible explanation is a premature rapid elimina-
tion of the low molecular weight derivative from the body
through glomerular filtration, which results in low therapeu-
tic concentration of the drug in the tumour tissue. Therefore,
attaching this oligopeptide derivative of PDM to a polymer
carrier is a plausible approach to improve the efficiency.
In this work, poly-[N-(2-hydroxyethyl)-l-glutamine]
(PHEG) and poly(ethylene glycol) (PEG)-grafted PHEG
(Fig. 2) were selected as drug carriers for the preparation
of the macromolecular prodrugs. PHEG is a synthetic poly-
␣-aminoacid, introduced by the group of Neri as a plasma
expander (Gerola et al., 1970). The lack of toxicity, its low im-
munogenicity, water solubility, biocompatibility, biodegrad-
ability and the presence of hydroxyl groups makes PHEG
a suitable carrier (Pytela et al., 1989, 1994, 1998; Rypacek
et al., 1985). In our research group, PDM and mitomycin C
(MMC) conjugates of PHEG were already synthetised and
evaluated (De Marre et al., 1994, 1995; Soyez and Schacht,
1997; Soyez et al., 1996a,b, 1997, 1999; Hoste et al., 2000,
2002). Due to the poor water solubility of hydrophobic drugs
like MMC and PDM, soluble conjugates are only obtained
for low degree of substitution. In order to maintain the wa-
ter solubility of the conjugates and to increase the degree of
substitution of the drug, PEG was used as solubilising entity.
It was shown in the work of Hoste et al. that PEG-grafted
dextran and PEG-grafted PHEG are promising candidates as
macromolecular drug carriers (Hoste et al., 1994, 1996, 2000;
Schacht and Hoste, 1997). The substitution of PEG onto the
PHEG resulted in a better solubility of the conjugates. More-
over, the in vivo evaluation of MMC-PHEG-PEG conjugates
showed a strong tumour regression within colorectal SW380
bearing mice (Hoste et al., 2002). Therefore, it was decided
to synthesise PDM conjugates with PEG-substituted PHEG
as carrier.
Fig. 1. Structure of N,N-di(2-chloroethyl)-4-phenylenediamine mustard
(PDM).
investigated by several groups (Duncan, 1992, 2003;
Kopecek et al., 2000; Subr et al., 1992; Rejmanova et al.,
1983; Soyez et al., 1996a,b; Matthews et al., 1996).
N,N-Di(2-chloroethyl)-4-phenylenediamine
mustard
(PDM; Fig. 1) is an alkylating antitumour agent which was
selected in this work because of its high cytotoxicity. Due to
the formation of an aziridium ion, it is highly reactive against
nucleophiles (Connors and Knox, 1995). The biological
activity of PDM is a result of the crosslinking of DNA
in the nucleus of the cell (Panthananickal et al., 1978;
Connors and Knox, 1995; Palmer et al., 1990). For aromatic
nitrogen mustards the substituent in para position has a
major influence on the chemical stability and reactivity of
the mustard. In case of PDM, where the para substituent is
an amine, the reactive aziridinium ion is stabilised, resulting
in a fast hydrolysis or a fast rate of alkylation. By acylation
of the aniline function of the nitrogen mustard the highly
toxic agent can be converted into a less toxic compound
(Palmer et al., 1992; Ross, 1955; Johnson and Owen, 1961).
Marquisee and coworkers developed nitrogen mustard
peptide derivatives such as Z-l-pro-l-leu-gly-l-pro-gly-
PDM which were designed to serve as a substrate for tumour-
associated collagenase (Marquisee and Kauer, 1978). The in
vitro results showed to be promising. However, the in vivo
tests using Sarcoma 180 cells were disappointing. This in
vivo lack of specific antitumour activity may be due to com-
peting cleavage by collagenases in certain normal tissues.
Fig. 2. Structure of poly-[N-(2-hydroxyethyl)-l-glutamine] (PHEG) and poly(ethylene glycol)-grafted PHEG (PEG-PHEG).

K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
161
This paper describes the synthesis and in vitro evaluation
of the PHEG and PEG-grafted PHEG conjugates of PDM
having a collagenase-sensitive oligopeptide spacer, l-pro-l-
leu-gly-l-pro-gly.
2.2.3. Coupling of PDM to Z-protected l-pro-gly
A solution of 1 g (3.28 mmol) Z-l-pro-gly 2 and 362 ␮l
(3.28 mmol) N-methylmorpholine in 10 ml dry tetrahydrofu-
ran was cooled to −10 ◦C. Four hundred and thirty-six mi-
croliters of isobutylchloroformiate (3.28 mmol) was added.
After reaction of 10 min at −10 ◦C the reaction mixture
was poured into 300 ml of a deaerated water–ice mixture.
The oil and white precipitation was centrifugated and sepa-
rated from the water phase. After washing the product with
deaerated water, the crude product was dissolved in chlo-
roform. The chloroform layer was dried over MgSO4 and
after filtration of the MgSO4 the solvent was evaporated.
Z-l-pro-gly-PDM 3 was purified by column chromatogra-
phy on silica with ethylacetate as eluent (Rf = 0.26). The re-
action product was characterised by 1H NMR spectroscopy
in CDCl3: δ = 1.80–2.20 ppm (4H, m, CH2CH2 l-pro-ring),
δ = 3.55 ppm (4H, t, N-CH2), δ = 3.65 ppm (4H, t, CH2Cl),
δ = 4.00–4.10 ppm (2H, ddd, ␣-CH2 gly), δ = 4.70 ppm (1H,
t, ␣-CH l-pro), δ = 5.10–5.20 ppm (2H, dd, CH2 ben-
zyl), δ = 6.60 ppm (2H, d, HA PDM), δ = 7.50 ppm (2H,
d, HB PDM), δ = 7.00–7.20 ppm (1H, bs, CONH PDM),
δ = 7.30 ppm (5H, bs, phenyl), δ = 8.40–8.80 ppm (1H, bs, NH
gly). Typical yield for the reaction: 70–75%.
2. Materials and methods
2.1. Chemicals
Aminoacids and peptides were obtained from Bachem
Chem. Co. (Bubendorf Switzerland). ␣-Hydroxy-␻-
methoxy-poly(ethylene glycol) of molecular weight 5000
Da, cathepsin B from bovine spleen (EC 3.4.22.1) and
cathepsin D from bovine spleen (EC 3.4.23.5), reduced
glutathione, ethylenediaminetetraacetic acid (EDTA) were
purchased from Fluka (Brussels, Belgium). Collagenase
type IV (Clostridiopeptidase A; EC 3.4.24.3) and human
serum albumin was obtained from Sigma Chem. Co. (St.
Louis, USA). All other chemicals were purchased from
Acros (Beerse, Belgium).
2.2. Synthesis of oligopeptide derivatives of PDM
2.2.1. Synthesis of PDM
2.2.4. Removal of Z-protecting group with HBr
1.15 g (2.14 mmol) Z-l-pro-gly-PDM 3 was suspended in
8 ml glacial acetic acid. The flask was fitted with a CaCl2
filled drying tube. Five millilitres of a 33% solution of HBr
in acetic acid was added. When the product was completely
dissolved (15–20 min), the reaction mixture was precipi-
tated in 150 ml dry ether. The precipitate was collected on
a filter, washed with ether and dried over NaOH pellets
in vacuo. The product was dissolved in superdry ethanol
(distilled with CaO, distilled with Mg, stored on molecular
N,N-Di(2-chloroethyl)-4-phenylenediamine
mustard
(PDM) was prepared according to the method of Soyez
et al. (1997). The monohydrochloride salt of PDM was char-
1
acterised by H NMR spectroscopy in D2O: δ = 7.30 ppm
(2H, d, HA phenyl); δ = 7.05 ppm (2H, d, HB phenyl),
δ = 3.90 ppm (4H, t, 2CH2Cl) and δ = 3.70 ppm (4H, t,
2NCH2).
For the analysis of PDM, an alkaline hydrolysis was
performed. The concentration of free chloride ions in so-
lution was measured using an Orion 9617 ionplus chlo-
ride electrode. For calibration standard chloride solutions
(10−4–10−2 M) were used.
˚
sieves 4 A) and reprecipitated in ether. The product was dried
over P2O5 and characterised by thin layer chromatography
1
(TLC; eluent: EtAc, Rf = 0) and by H NMR spectroscopy
in CD3OD: δ = 2.00–2.50 ppm (4H, m, CH2CH2 l-pro-ring),
δ = 3.30–3.40 ppm (2H, m, N-CH2 l-pro), δ = 3.65 ppm (4H,
t, NCH2 PDM), δ = 3.75 ppm (4H, t, CH2Cl), δ = 4.10 ppm
(2H, dd, ␣-CH2 gly), δ = 4.40 ppm (1H, t, ␣-CH l-pro),
δ = 6.90 ppm (2H, d, HA PDM), δ = 7.50 ppm (2H, d, HB
PDM). Yield: 70–80%.
2.2.2. Synthesis of Z-protected l-pro-gly 2
To a solution of 300 mg (1.76 mmol) l-pro-gly 1 and
470 mg K2CO3 in 5 ml milliQ water, 575 ␮l benzylchloro-
formiate was added. After reaction overnight the mixture
was extracted three times with 5 ml diethylether. The aque-
ous layer was acidified to pH 2 with 1N HCl and was ex-
tracted with ethylacetate. The solvent was evaporated and
the product Z-l-pro-gly was dried in vacuo. The reaction
product was characterised by 1H NMR spectroscopy in
CD3OD: δ = 1.90–2.30 ppm (4H, m, CH2CH2 l-pro-ring),
δ = 3.55 ppm (4H, t, NCH2), δ = 3.65 ppm (4H, t, CH2Cl),
δ = 3.90–4.20 ppm (2H, ddd, ␣-CH2 gly), δ = 4.40 ppm (1H,
t, ␣-CH l-pro), δ = 5.10–5.20 ppm (2H, dd, CH2 benzyl),
δ = 6.60 ppm (2H, d, HA PDM), δ = 6.80 ppm (1H, bs, CONH
PDM), δ = 7.30 ppm(5H, bs, phenyl), δ = 7.50 ppm(2H, d, HB
PDM), δ = 8.00–8.40 ppm (1H, bs, NH gly). Typical yield for
the reaction: 85–90%.
2.2.5. Synthesis of N-hydroxysuccinimidyl ester of
Z-l-pro-l-leu-gly
A solution of 780 mg (1.85 mmol) Z-protected l-pro-l-
leu-gly 5 and 220 mg (1.85 mmol) N-hydroxysuccinimide in
10 ml dry chloroform was cooled to 0 ◦C. A cooled solution
of 386 mg dicyclohexylcarbodiimide (DCC; 1.85 mmol) in
chloroform was added. The mixture was stirred overnight at
0 ◦C. After filtration of dicyclohexylurea (DCU), the solvent
was evaporated and the product was redissolved in chloro-
form. The solution was again filtered. After evaporation of
the solvent, the reactive ester 6 was obtained. The ester was
characterised by IR spectroscopy (NHS-ester: 1738 cm−1
)

162
K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
1
2.3. Synthesis of macromolecular PDM conjugates
and by H NMR spectroscopy: δ = 0.90 ppm (6H, m, CH3
l-leu), δ = 1.40–1.80 ppm (4H, m, CH2CH2 l-pro-ring),
δ = 1.90–2.30 ppm (3H, m, CHCH2 l-leu), δ = 2.80 ppm
(4H, bs, CH2CH2NHS-ester), δ = 3.30–3.60 ppm (2H, m,
N-CH2 l-pro), δ = 4.10–4.20 ppm (2H, ddd, ␣-CH2 gly),
δ = 4.35 ppm (1H, m, ␣-CH of l-leu), δ = 4.50 ppm (1H, t, ␣-
CH l-pro), δ = 5.20 ppm (2H, bs, CH2 benzyl), δ = 6.50 ppm
(1H, d, NH l-leu), δ = 6.70 ppm (1H, t, NH gly) and
δ = 7.50 ppm (5H, bs, phenyl).
A solution of 100 mg of PHEG in 12.5 ml DMSO and
7.5 ml pyridine (5/3) was cooled to 0 ◦C. One hundred
and eight milligram (0.7 eq.) N,N-disuccinimidylcarbonate
(DSC) was added. After 1.5 h reaction at 0 ◦C the reaction
mixture was precipitated in isopropanol. After washing, the
DSC activated PHEG was dissolved in 10 ml DMSO and
15 ml pyridine. Thirty-five milligram of oligopeptide-PDM
4 and a few drops of N-methylmorpholine were added. In case
of the PEG-grafted PHEG, the reaction was stirred overnight
atroomtemperature, afterwhich88 mgPEG-NH2 wasadded.
After 48 h the conjugate was precipitated in isopropanol. The
product was washed and dried. Finally, the conjugate was
purified by preparative GPC (Sephadex G25) with water as
eluent. The conjugate was freezedried. The degree of sub-
stitution was determined by UV analysis (λM = 272 nm and
εM = 17,390 l mol−1cm−1) and by 1H NMR spectroscopy in
D2O: δ = 6.9 ppm and 7.5 ppm: aromatic protons of PDM.
2.2.6. Synthesis of Z-protected oligopeptide-PDM
derivatives
For the preparation of Z-l-pro-l-leu-gly-l-pro-gly-PDM
7 636 mg (1.23 mmol) NHS-ester 6 was dissolved in 20 ml
1,2-dimethoxymethane and cooled to 0 ◦C. Five hundred
and seventy-seven milligram (1.23 mmol) l-pro-gly-
PDM·HBr and 103 mg (1.23 mmol) NaHCO3 (1.23 mmol)
was dissolved in 1 ml deaerated water. After adding 3 ml
1,2-dimethoxyethane to the l-pro-gly-PDM 2 solution, the
mixture was cooled to 0 ◦C. After precipitation in 500 ml
of a deaerated water–ice mixture, the product was isolated
after centrifugation and washed with deaerated water. The
solid was dissolved in chloroform and dried on MgSO4.
After evaporation of the solvent, the product was purified by
recristallisation from ethylacetate/cyclohexane. Yield: 40%.
The product was characterised by TLC (eluent: EtAc,
Rf = 0.15) and by 1H NMR spectroscopy in CDCl3:
δ = 0.80 ppm (6H, m, CH3 l-leu), δ = 1.10–1.30 ppm
(3H, m, CHCH2 l-leu), δ = 1.80–2.30 ppm (8H, m,
2CH2CH2 l-pro-ring), δ = 3.30–3.70 ppm (12H, m, N-CH2
l-pro + NCH2CH2Cl PDM), δ = 4.15 ppm (2H, dd, ␣-CH2
gly), δ = 4.25 ppm (1H, m, ␣-CH2 l-leu), δ = 4.35 ppm (1H,
dd, ␣-CH2 gly), δ = 4.50 ppm (1H, t, ␣-CH l-pro next to Z-
group), δ = 4.60 ppm (1H, t, ␣-CH l-pro), δ = 6.90 ppm (2H,
d, HA PDM) and 7.50 ppm (2H, d, HB PDM).
2.4. In vitro hydrolytic stability of macromolecular
PDM conjugates
The hydrolysis of the nitrogen mustard in the polymeric
prodrugs was investigated at 20 ◦C in phosphate buffer pH
7.4 (0.01 M KH2PO4) and citrate buffer pH 5.5 (0.01 M
Na2HPO4/citric acid) using a chloride ion selective electrode.
Four milligrams of the conjugate was dissolved in 4 ml of
buffer. Eighty microliters of 1N NaNO3 solution was added
as ionic strength adjustor (ISA). The concentration of free
chloride ions in solution was measured after calibration with
standard chloride solutions (10−4–10−2 M) using an Orion
9617 ionplus chloride electrode.
2.5. In vitro enzymatic degradation of macromolecular
PDM conjugates
The synthesis of Z-GABA-l-pro-l-leu-gly-l-pro-gly-
PDM 9 was established as described for 8 with Z-GABA-
NHS and l-pro-l-leu-gly-l-pro-gly-PDM instead of prod-
ucts 6 and 4.
2.5.1. Degradation in presence of bacterial collagenase
type IV
The release of PDM in presence of collagenase clostrid-
ium histolyticum was studied in phosphate buffered saline
(PBS) buffer pH 7.4 at 37 ◦C. One milligram of conjugate
was dissolved in 600 ␮l of phosphate buffer. Four hundred
microliters of collagenase solution (2 mg/ml) was added. At
regular times samples were taken and analysed by high per-
formance liquid chromatography (HPLC) (column: Zorbax
ODS 4.6 mm × 150 mm, flow: 1 ml/min, UV detection at
265 nm) using a 35/65 acetonitrile/acetic acid (0.05 M, pH
4) mixture. Calibration of the released products was carried
out using PDM, gly-PDM and gly-l-pro-gly-PDM solutions
of known concentrations.
2.2.7. Synthesis of oligopeptide-PDM derivatives
The synthesis of l-pro-l-leu-gly-l-pro-gly-PDM 8
and GABA-l-pro-l-leu-gly-l-pro-gly-PDM 10 was accom-
plished by removal of the Z-group with a 33% HBr so-
lution in acetic acid as described above for the prepara-
tion of product 4. The peptide-PDM derivative was charac-
1
terised by H NMR spectroscopy in CD3OD: δ = 0.90 ppm
(6H, m, CH3 leu), δ = 1.45–1.75 ppm (3H, m, CHCH2
leu), δ = 1.90–2.50 ppm (10H, m, 2CH2CH2 pro-ring + NH-
CH2 pro-ring), δ = 3.55–3.80 ppm (12H, m, N-CH2 pro-
ring + NCH2CH2Cl), δ = 3.90–4.15 ppm (4H, m, 2 ␣-CH
gly), δ = 4.30 ppm (1H, t, ␣-CH pro between 2 gly),
δ = 4.40 ppm (1H, m, ␣-CH leu), δ = 4.45 ppm (1H, t, ␣-CH
pro), δ = 6.80 ppm (2H, d, HA PDM), δ = 7.50 ppm (2H, d,
HB PDM).
2.5.2. Degradation in presence of cathepsin B
The degradation of the conjugates, in presence of cathep-
sin B from bovine spleen, was studied at 37 ◦C in buffer pH
5.5 (0.2 M Na2HPO4/citric acid). One milligram of conjugate

K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
163
was dissolved in 750 ␮l phosphate buffer (pH 5.5), 100 ␮l of
a 10-mM EDTA solution, 100 ␮l of a 50-mM reduced glu-
tathione solution and 50 ␮l of the cathepsin B stock solution
(2.5 mg/ml) were added. Samples were taken at appropriate
time intervals and analysed by HPLC as decribed above.
2.5.3. Degradation in presence of cathepsin D
The degradation of the conjugates in presence of cathepsin
D from bovine spleen was studied at 37 ◦C in buffer pH 3. One
milligram of conjugate was dissolved in 875 ␮l of buffer pH
3 (0.1 M citrate buffer) and 125 ␮l of the cathepsin D solution
(1 and 2 mg/ml) was added. Samples were taken after regular
time intervals and analysed by HPLC as described above.
2.5.4. Degradation in presence of tritosomes
The degradation of the conjugates in presence of isolated
rat liver lysosomal enzymes (tritosomes) was studied at 37 ◦C
in buffer pH 5.5. One milligram of conjugate was dissolved in
400 ␮l phosphate buffer (0.2 M Na2HPO4/citric acid, 0.2%
triton X-100, w/v), 100 ␮l of a 10-mM EDTA solution, 100 ␮l
of a 50-mM reduced glutathione solution and 400 ␮l trito-
somes were added. Samples were taken at appropriate time
intervals and analysed by HPLC as decribed before.
Fig. 3. Synthesis of the PEG-grafted PHEG-PDM conjugate.
performed using water as a solvent at 25 ◦C and in absence
as well as in presence of human serum albumin.
2.7. In vitro evaluation against tumour cells
HT1080 fibrosarcoma cells or MDA-MB 231 adenocar-
cinoma cells were plated at a density of 104 cells/well in
96-well plates and maintained at 37 ◦C in culture medium
for 24 h. Both cell lines are derived of human epithelial cells
of the breast. HT1080 cells were maintained in RPMI media
with foetal calf serum (5%, v/v) and glutamine (1%, v/v).
2.6. Dynamic laser light scattering studies
Dynamic laser light scattering (DLS) studies were per-
formed with a DLS-700 (Polymer Laboratories) spectrome-
ter, using an Ar laser (λ = 488 nm). The measurements were
Fig. 4. Reaction scheme for the synthesis of the l-pro-gly-PDM 4.

164
K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
Fig. 5. Reaction scheme for the synthesis of l-pro-l-leu-gly-l-pro-gly-PDM 8.
MDA cells were maintained in DME media with foetal calf
serum (10%, v/v) and glutamine (1%, v/v). The PDM conju-
gates were added at various concentrations and left for 72 h.
Cell viability was assessed after 72 h by measuring the cells
ability to reduce the tetrazolium dye 3-(4,5-dimethyl)thiazol-
2,5-diphenyl tetrazolium bromide (MTT; Sgouras and
Duncan, 1990). The results were expressed as the percent-
age of viability of cells in control wells.
tion by benzyloxycarbonyl (Z)-group is well established in
peptide chemistry (Bodansky and Bodansky, 1984). The Z-
group was introduced by reaction of the dipeptide l-pro-gly
with benzylchloroformiate.
The coupling of the peptide fragments to the nitogen
mustard can be readily accomplished using the mixed
anhydride method (Marquisee and Kauer, 1978; Soyez
et al., 1997). Thus, in a second step, the Z-protected dipep-
tide 2 was coupled to PDM. Using isobutylchloroformiate
and N-methylmorpholine to form a mixed anhydride of 2.
The latter was subsequently reacted with the aniline function
of the nitrogen mustard.
3. Results and discussion
The removal of the Z-group was achieved by treatment of
Z-l-pro-gly-PDM 3 with a 33% HBr solution in acetic acid
(AcOH) leading to the dipeptide-PDM 4.
The pentapeptide-PDM fragment was synthetised follow-
ing the strategy presented in Fig. 5.
The reactive ester of the commercially available Z-l-
pro-l-leu-gly was prepared by dicyclohexylcarbodiimide
(DCC) coupling of the Z-protected tripeptide 5 and
N-hydroxysuccinimide (NHS).
3.1. Synthesis of polymeric PDM conjugates
The general strategy for the synthesis of the conjugates
is shown in Fig. 3. The selected carrier, PHEG, is acti-
vated by reaction with N,N-disuccinimidylcarbonate. The
oligopeptide-PDM moieties and PEG-NH2 are coupled onto
the polymer via their amino endgroups.
3.1.1. Synthesis of oligopeptide-PDM
The coupling of the dipeptide-PDM 4 to the NHS-
ester 5 was accomplished in 1,2-dimethoxymethane as sol-
vent.
In an attempt to achieve a better coupling efficiency
with the polymeric carrier, GABA-l-pro-l-leu-gly-l-pro-
gly-PDM 10 with ␥-aminobutyric acid (GABA) was syn-
Inordertosynthesisethel-pro-l-leu-gly-l-pro-gly-PDM,
the alkylating drug PDM is coupled with an amine protected
dipeptide. The synthesis of the dipeptide-PDM derivative is
shown in Fig. 4.
The first step in the synthesis of l-pro-gly-PDM involves
the protection of the amino function of the dipeptide. Protec-

K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
165
Table 2
Size of PHEG and the PHEG-PDM conjugates
Diameter (nm)a
d1
d2
d3
PHEG81000
PHEG-l-pro-l-leu-gly-l-pro-
gly-PDM
PHEG-GABA-l-pro-l-leu-gly-l-
pro-gly-PDM
PHEG-PEG-l-pro-l-leu-gly-l-
pro-gly-PDM
–
–
15.6
18.6
–
143.6
146.5
116.8
–
–
23.1
20.8
Fig. 6. Reaction scheme for the synthesis of GABA-l-pro-l-leu-gly-l-pro-
gly-PDM 10.
Human serum albumin (HSA)
PHEG 81000 + HSA
PHEG-l-pro-l-leu-gly-l-pro-
gly-PDM + HSA
7.5
6.3
5.7
–
15.2
16.9
151.0
107.5
189.1
thetised (Fig. 6). Unfortunately, a higher degree of substitu-
tion could not be achieved.
3.1.2. Synthesis of the PDM conjugates
PHEG-GABA-l-pro-l-leu-gly-l-
pro-gly-PDM + HSA
a
6.8
15.7
228.1
Aiming to couple to the oligopeptide-PDM moieties and
PEG-NH2 ontothepolymericcarrierPHEG(Mw = 106 kDa),
the hydroxyl groups should be activated. Reaction with
N,N-disuccinimidylcarbonate (DSC) in DMSO/pyridine
(5/3) resulted in a partial conversion of the hydroxyl groups
into reactive carbonate groups (Fig. 3).
d1–d3 indicate the average diameter of the different distributions of the
polymer and conjugates.
tions, whereas larger particles and aggregates generate slower
fluctuations. The rate of fluctuations can be determined by au-
tocorrelation analysis. This allows to calculate the diffusion
coefficient, D, and particle sizes which are related via the
Stokes–Einstein equation (Berne and Pecora, 1976).
The activated polymer and the oligopeptide-PDM deriva-
tive were reacted in a mixture of DMSO/pyridine (3/2) for
48 h. In case of the PEG-substituted PHEG-PDM conjugate,
the activated PHEG was reacted first with the oligopeptide-
PDM moiety. After 24 h PEG-NH2 was added to the reaction
mixture. The conjugate was isolated by precipation in iso-
propanol and further purification was performed by prepara-
tive GPC (Sephadex G25). The degree of PDM substitution
kT
D =
6πηrh
where k is the Boltzmann constant; T is the absolute temper-
ature; η is the viscosity of the solvent; and rh is the hydrody-
namic radius.
For the PDM conjugates, a bimodal distribution was ob-
served (Table 2). The smallest particles are the free polymer
chains. The larger particles are formed by aggregation. How-
ever, the weight and number percentage of the aggregates is
small.
1
was determined by UV spectrometry at 272 nm and by H
NMR spectroscopy.
The molar substitution degree of PDM and PEG for the
different PDM conjugates are given in Table 1. As mentioned
before, the introduction of a GABA-containing spacer did not
result in a better coupling yield.
When a conjugate is administered intravenously, it is ex-
posed to blood proteins such as serum albumine. In order to
examine a possible effect of serum albumin on the particle
size of the conjugates DLS measurements were performed in
presence of human serum albumin. The size of human serum
albumin is approximately 6–8 nm. The addition of albumin
did not alter the size of the other particle distribution (d2 and
d3) significantly (population with lowest diameter, d1).
3.2. Dynamic light scattering measurements
The coupling of a hydrophobic drug such as PDM onto a
polymeric carrier may result in an aggregation of the resulting
conjugate. Therefore, the size of the polymeric prodrugs was
determined by dynamic laser light scattering (DLS). DLS is
based on the scattering of light by diffusing particles. Due
to the Brownian motion of the molecules in the solution,
the intensity of the light fluctuates as a function of time.
Small rapidly diffusing particles will result in fast fluctua-
3.3. In vitro hydrolytic stabilility of polymeric PDM
derivatives
Table 1
The hydrolytic stability of the chloroethyl groups of PDM
in the conjugates was studied using a chloride selective elec-
trode. ThePDMconjugateswereincubatedat20 ◦Cindiluted
phosphate(pH7.4), respectively, citrate(pH5.5)buffers, sim-
ulating the physiological and lysosomal pH, respectively. The
concentration of chloride ions the solution was measured at
regular time intervals with a chloride selective electrode cal-
ibrated with standard potassium chloride solutions. The per-
centage hydrolysis of the chloroethyl groups was calculated
Composition of a series of PHEG-PDM conjugates
PDM conjugate
mol% in 1H NMR
mol% in UV
2.3
PHEG-l-pro-l-leu-gly-l-pro-
gly-PDM
2.4
PHEG-GABA-l-pro-l-leu-gly-l-
pro-gly-PDM
PHEG-l-pro-l-leu-gly-l-pro-
gly-PDM
1.8
3.2
4.2
1.8
3.1
–
PHEG-PEG

166
K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
The work of Marquisee demonstrated that Z-l-pro-l-leu-
gly-l-pro-gly-PDM is a good substrate for tumour-associated
and bacterial collagenase (Marquisee and Kauer, 1978).
Due to high cost of human MMPs, the bacterial collage-
nase, which is known to degrade collagen, was used as model
in this work.
To examine whether the polymeric PDM derivatives are
still usefull delivery systems of PDM, degradation studies
were performed in presence of collagenase clostridium his-
tolyticum, tritosomes (lysosomal enzymes isolated from liver
of rats), cathepsin B and D.
The concentration of PDM and gly-l-pro-gly-PDM in so-
lution was calculated from the peak areas of the HPLC chro-
matograms after calibration with PDM and gly-pro-gly-PDM
standards.
The release of PDM and gly-pro-gly-PDM of the macro-
molecular PDM prodrugs in presence of collagenase clostrid-
ium histolyticum are presented in Fig. 8a and b.
Fig. 7. Hydrolytic stability of PDM and PDM conjugates.
by comparing the measured chloride concentrations during
hydrolysis with the maximum chloride concentration calcu-
lated based on the degree of substitution.
As previously reported (Soyez et al., 1997), the hydrolytic
stability of PDM, when covalently attached to the polymeric
carrier, results in a strong increase in stability.
The percentage unhydrolysed chloride of free PDM and
PDM conjugates as function of time is represented in Fig. 7.
The same change in hydrolytic stability and reactivity was
also observed in this work. The PEG side chains have no
significant effect on the hydrolysis of the macromolecular
PDM derivative.
3.4. In vitro enzymatic degradation of polymeric PDM
derivatives
Ideally, the macromolecular antitumour prodrug should
selectively deliver the cytotoxic compound in or near the
tumour site. It has been demonstrated that tumour cells are
able to invade the normal tissue due to a high expression
of proteases. Enzymes such as matrix metalloproteases
(MMPs: collagenase, stromelysines) (Stetler-Stevenson,
1990; Gervasi et al., 1996; Young et al., 1996; McMillan
et al., 1996; Ginestra et al., 1997; Moore et al., 1997; von
Bredow et al., 1998), cathepsines (B and D) (Ferguson
et al., 1973; Laurent-Martha et al., 1998; Kazakova and
Orekhovich, 1972; Rochefort et al., 1990) and plasminogen
activators (Moser et al., 1994) play an important role in the
degradation of the basal membrane and the extracellular
matrix of healthy cells. The MMPs are most involved in
the metastatic and invassive process and exhibit a high
specificity for type IV collagen (Stetler-Stevenson, 1990),
that makes up the backbone of the basal membrane. The
proteolytic activity of these tumour-associated collagenases
can be explored in the concept of macromolecular prodrugs
for releasing the anticancer agent in or near the tumour site.
Fig. 8. (a) Enzymatic degradation of the macromolecular PDM conjugates
in presence of collagenase clostridium histolyticum: release of PDM and
gly-l-pro-gly-PDM as function of time. (b) Enzymatic degradation of the
macromolecular PDM conjugates in presence of collagenase clostridium
histolyticum: total amount of PDM (=free PDM and gly-l-pro-gly-PDM)
released as function of time.

K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
167
The observed lack of in vitro antitumour activity is most
likely due to the hydrolytical instability of the drug moiety.
4. Conclusions
Macromolecular PDM prodrugs with a collagenase-
sensitive oligopeptide spacer, l-pro-l-leu-gly-l-pro-gly,
were synthetised. Conjugates based on PHEG and PEG-
grafted PHEG carriers form aggregates in water. The size of
their aggregates is not altered in presence of serum albumin.
In vitro degradation studies in presence of lysosomal and
tumour-associated enzymes showed that the conjugates are
specific substrates for collagenase. No antitumour activity
was shown in vitro, which might be due to the hydrolytical
instability of the chloride groups in the drug moiety. Despite
the negative in vitro antitumour effect, the concept of linking
antitumour drugs via a collagenase-sensitive spacer onto
PHEG can be succesful for more hydrolysis-stable drugs
like mitomycin C. The synthesis and in vitro drug release of
such conjugates are discussed in a forthcoming paper.
Fig. 9. In vitro cytotoxicity of macromolecular PDM conjugates against
HT1080 cell line.
The conjugates did not exclusively release the alkylat-
ing PDM. The peptidyl-PDM fragment, gly-l-pro-gly-PDM,
was also released from the macromolecular derivatives. The
decrease in PDM concentration at higher incubation times
(above 60 min) can be attributed to the hydrolysis of the PDM
chloroethyl groups.
Adding GABA to the peptidyl-spacer gives a slightly
faster release of gly-l-pro-gly-PDM. This is probably due
to a better accessibility of the enzyme to the l-leu-gly bond.
Attachment of PEG sidechains onto the polymer backbone
results in a slightly slower release.
Acknowledgements
Although all conjugates proved to be good substrates for
collagenase, no release of PDM nor peptidyl-PDM fragments
was observed in presence of tritosomes (lysosomes isolated
from liver of rats), cathepsin B and D (data not shown). This
data indicates that a l-pro-l-leu-gly-l-pro-gly spacer may be
a suitable choice for achieving PDM release by collagenases.
The authors would like to thank the Flemish institute
for the promotion of Scientific-Technological Research in
Industry (IWT), the Fund for Scientific Research-Flanders
(FWO) and the Belgian Ministry of Scientific Programming,
IUAP/PAI-V/03.
3.5. In vitro evaluation against tumour cells
References
The in vitro cytotoxicity of the polymeric PDM prodrugs
was evaluated against HT1080 fibrosarcoma cells and MDA-
MB 231 adenocarcinoma cells. These cancer cell lines were
selected because they secrete, respectively, high and low con-
centrations of type IV collagenase (Ginestra et al., 1997;
Moore et al., 1997; Young et al., 1996).
These preliminary experiments show that the PDM con-
jugates with PHEG as carrier are not cytotoxic in the concen-
tration range used over a period of 72 h (Figs. 9 and 10).
Bodansky, M., Bodansky, A., 1984. The Practice of Peptide Synthesis.
Springer-Verlag, Berlin, Heidelberg, New York, Tokyo, pp. 12–15.
Berne, B., Pecora, R., 1976. Dynamic Light Scattering and Applications
to Chemistry, Biology and Physics. Wiley Interscience.
Connors, T.A., Knox, R.J., 1995. Prodrugs in cancer-chemotherapy. Stem
Cells 13, 501–511.
De Marre, A., Soyez, H., Schacht, E.H., 1994. Synthesis of macromolec-
ular Mitomycin C derivatives. J. Control. Release 32, 129–137.
De Marre, A., Soyez, H., Schacht, E.H., Shoaibi, M.A., Seymour, L.W.,
Rihova, B., 1995. Synthesis and evaluation of macromolecular pro-
drugs of mitomycin C. J. Control. Release 36, 87–97.
Duncan, R., 1992. Drug–polymer conjugates: potential for improved
chemotherapy. Anticancer Drugs 3, 175–210.
Duncan, R., 2003. The dawning era of polymer therapeutics. Nat. Rev.
Drug Discov. 2, 347–360.
Ferguson, J.B., Andrews, J.R., Voynick, I.M., Fruton, J.S., 1973. The
specificity of cathepsin D. J. Biol. Chem. 19, 6701–6708.
Gerola, A., Atoni, G., Benvenuti, F., Cocola, F., Neri, P., 1970. Poly-
N-[(2-hydroxyethyl)-l-glutamine], a new plasma expander. Adv. Exp.
Med. Biol. 9, 329–338.
Gervasi, D.C., Raz, A., Dehem, M., Yang, M., Kurkinen, M., Fridman, R.,
1996. Carbohydrate-mediated regulation of matrix metalloproteinase-
2 activation in normal human fibroblast and fibrosarcoma cells.
Biochem. Biophys. Res. Commun. 228, 530–538.
Fig. 10. In vitro cytotoxicity of macromolecular PDM conjugates against
MDA-MB 231 cell line.
Ginestra, A., Monea, S., Seghezzi, G., Dolo, V., Nagase, H., Mignatti,
P., Vittorelli, M., 1997. Urokinase plasminogen activator and gelati-

168
K. De Winne et al. / European Journal of Pharmaceutical Sciences 24 (2005) 159–168
nases are associated with membrane vesicles shed by human HT1080
fibrosarcoma cells. J. Biol. Chem. 272, 17216–17222.
Hoste, K., Bruneel, D., De Marre, A., De Schrijver, F., Schacht, E.,
1994. Synthesis and characterization of poly(oxyethylene) modified
dextrans. Macromol. Rapid Commun. 15, 697–704.
Hoste, K., Bruneel, D., De Marre, A., De Schrijver, F., Schacht, E.,
1996. Synthesis, characterization, and in vitro biodegradation of
poly(ethylene glycol) modified poly[N5-(2-hydroxyethyl-l-glutamine].
J. Bioact. Compat. Pol. 11, 85–99.
Hoste, K., Seymour, L.W., Schacht, E.H., 2000. New derivatives of polyg-
lutamic acid as drug carrier systems. J. Control. Release 64, 53–61.
Hoste, K., Schacht, E., Rihova, B., 2002. Synthesis and biological eval-
uation of PEG-substituted macromolecular prodrugs of mitomycin C.
J. Bioact. Compat. Pol. 17, 123–138.
Johnson, B.J., Owen, L.N., 1961. Cytotoxic compounds. Part II:
some Nꢀ[␣-(aminoacyl]-N,N-di-2ꢀ-chloroethyl-p-phenylenediamines. J.
Chem. Soc., 3401–3404.
Maeda, H., Matsumura, Y., 1989. Tumoritropic and lymphotropic princi-
ples of macromolecular drugs. Crit. Rev. Ther. Drug 6, 193–210.
Matthews, S.E., Pouton, C.W., Threadgill, M.D., 1996. Macromolecu-
lar systems for chemotherapy and magnetic resonance imaging. Adv.
Drug Deliv. Rev. 18, 219–267.
Pytela, J., Saudek, V., Drobnik, J., Rypacek, F., 1989. Poly(N5-
hydroxyalkylglutamines). 4. Enzymatic degradation of N5-(2-
hydroxyethyl)-l-glutamine homopolymers and copolymers. J. Control.
Release 10, 17–25.
Pytela, J., Kotva, R., Rypacek, F., 1998. Enzymatic degrada-
tion of poly[N5-(2-hydrodryethyl)-l-glutamine] and poly[N5-(2-
hydroxyethyl)-l-glutamine-stat-l-glutamic acid]: analysis of final
degradation products. J. Bioact. Compat. Pol. 13, 198–209.
Rejmanova, P., Kopecek, J., Pohl, J., Baudys, M., Kostka, V., 1983.
Polymers containing enzymatically degradable bonds. 8. Degrada-
tion of oligopeptide sequences in N-(2-hydroxypropyl)methacrylamide
co-polymers by bovine spleen cathepsin B. Macromol. Chem. 184,
2009–2020.
Ringsdorf, H., 1975. Structure and properties of pharmacologically active
polymers. J. Polym. Sci. Polym. Symp. 51, 135–153.
Rochefort, H., Capony, F., Garcia, M., 1990. Cathepsin D: a protease in-
volved in breast cancer metastasis. Cancer Metastasis Rev. 9, 321–331.
Ross, W.C.J., 1955. The chemistry of cytotoxic alkylating agents. Adv.
Cancer Res. 1, 397–449.
Rypacek, F., Saudek, V., Pytela, J., Skarda, V., Drobnik, J., 1985.
Polyglutamines—versatile biodegradable carriers. Macromol. Chem.
Suppl. 9, 129–135.
Marquisee, M., Kauer, J., 1978. Collagenase-sensitive peptidyl-nitrogen
mustards as potential anti-tumor agents. J. Med. Chem. 12,
1188–1194.
McMillan, J.I., Weeks, R., West, J.W., Bursten, S., Rice, G.C., Lovett,
D.H., 1996. Pharmacological inhibition of gelatinase B induction and
tumor cell invasion. Int. J. Cancer 67, 523–531.
Moore, D.H., Allison, B., Look, K.Y., Sutton, G.P., Bigsby, R., 1997.
Collagenase expression in ovarian cancer cell lines. Gynecol. Oncol.
65, 78–82.
Schacht, E.H., Hoste, K., 1997. Poly(ethylene glycol)-grafted polymers
as drug carriers. Poly(ethylene glycol) ACS Symp. Ser. 680, 297–315.
Sgouras, D., Duncan, R., 1990. Methods for the evaluation of biocompati-
bity of soluble synthetic polymers which have potential for biomedical
use. 1. Use of tetrazolium-based colorimetric assay (MTT) as a pre-
liminary screen for evaluation of in vitro cytotoxicity. J. Mater. Sci.
Mater. Med. 1, 61.
Stetler-Stevenson, W.G., 1990. Typer IV collagenases in tumour invasion
and metastasis. Cancer Metastasis Rev. 9, 289–303.
Moser, T.L., Young, T.N., Rodriguez, G.C., Pizzo, S.V., Bast, R.C.,
Stack, M.S., 1994. Secretion of extracellular matrix-degrading pro-
teinases is increased in epithelial ovarian carcinoma. Int. J. Cancer 56,
552–559.
Kazakova, O.V., Orekhovich, V.N., 1972. Effect of cathepsins D from
normal and malignant tissues on synthetic peptides. J. Biol. Chem.
247, 4224–4228.
Kopecek, J., Kopeckova, P., Minko, T., Lu, Z.-R., 2000. HPMA
copolymer-anticancer drug conjugates: design, activity, and mecha-
nism of action. Eur. J. Pharm. Biopharm. 50, 61–81.
Laurent-Martha, V., Farnoud, M.R., Lucas, A., Rougeot, C., Garcia, M.,
Rochefort, H., 1998. Endocytosis of pro-cathepsin D into breast cancer
cells is mostly independent of mannose-6-phosphate receptors. J. Cell
Sci. 111, 2539–2549.
Palmer, B.D., Wilson, W.R., Cliffe, S., Denny, W.A., 1990. Hypoxia-
selective antitumour agents. 3. Relationships between structure and
cytotoxicity against cultured tumor cells for substituted N,N-bis(2-
chloroethyl)anilines. J. Med. Chem. 33, 112–121.
Palmer, B.D., Wilson, W.R., Cliffe, S., Denny, W.A., 1992. Hypoxia-
selective antitumor agents. 5. Synthesis of water-soluble nitroaniline
mustards with selective cytotoxicity for hypoxic mammalian-cells. J.
Med. Chem. 35, 3214–3222.
Panthananickal, A., Hansch, C., Leo, A., 1978. Structure-activity-
relationships in anti-tumor aniline mustards. J. Med. Chem. 21, 16–26.
Pytela, J., Jakes, J., Rypacek, F., 1994. A study of enzymatic degradation
of a macromolecular substrate, poly[N5-2-hydroxyethyl)-l-glutamine],
by gel-permeation chromatography and kinetic modeling. Int. J. Biol.
Macromol. 16, 15–20.
Soyez, H., De Marre, A., Schacht, E., 1996a. Polymeric prodrugs of
mitomycin C. Macromol. Symp. 103, 163–176.
Soyez, H., Schacht, E., Vanderkerken, S., 1996b. The crucial role of
spacer groups in macromolecular prodrug design. Adv. Drug Deliv.
Rev. 21, 81–106.
Soyez, H., Schacht, E., 1997. Macromolecular derivatives of N,N-di-(2-
chloroethyl)-4-phenylenediamine mustard. 1. Synthesis and in vitro
hydrolytic stability. J. Control. Release 45, 235–247.
Soyez, H., Schacht, E., Jelinkova, M., Rihova, B., 1997. Biological eval-
uation of mitomycin C bound to a biodegradable polymeric carrier. J.
Control. Release 47, 71–80.
Soyez, H., Seymour, L.W., Schacht, E., 1999. Macromolecular derivatives
of N,N-di-(2-chloroethyl)-4-phenylene diamine mustard. 2. In vitro
cytotoxicity and in vivo anticancer efficacy. J. Control. Release 57,
187–196.
Subr, V., Strohalm, J., Ulbrich, K., Duncan, R., Hume, I., 1992. Poly-
mers containing enzymatically degradable bonds. 12. Effect of spacer
structure on the rate of release of daunomycin and adriamycin from
poly[N-(2-hydroxypropyl)-methacrylamide] copolymer drug carriers
in vitro and antitumor-activity measured in vivo. J. Control. Release
18, 123–132.
von Bredow, D.C., Cress, A.E., Howard, E.W., Bowden, G.T., Nagle, R.B.,
1998. Activation of gelatinase-tissue-inhibitors of metalloproteinase
complexes by matrilysin. Biochem. J. 331, 965–972.
Young, T.N., Rodriguez, G.C., Rinehart, R., Bast, R.C., Pizzo, S.V., Stack,
M.S., 1996. Characterization of gelatinases linked to extracellular ma-
trix invasion in ovarian adenocarcinoma: Purification of matrix met-
alloproteinase 2. Gynecol. Oncol. 62, 89–99.