Poly(ethylene glycol) Transport Forms of Vancomycin: A Long-Lived Continuous Release Delivery System

Article abstract of DOI:10.1021/jm030202g

The facile reaction of vancomycin with various PEG linkers, at the V ₃ position, has been selectively accomplished by using an excess of base in DMF. Using rPEG as a blocking group for V₃ provides crystalline derivatives that can be

Full text of DOI:10.1021/jm030202g

J . Med. Chem. 2003, 46, 5021-5030
5021
P oly(eth ylen e glycol) Tr a n sp or t F or m s of Va n com ycin : A Lon g-Lived
Con tin u ou s Relea se Deliver y System
Richard B. Greenwald,* Hong Zhao, J ing Xia, and Anthony Martinez
Enzon Pharmaceuticals Inc., 20 Kingsbridge Road, Piscataway, New J ersey 00854
Received April 29, 2003
The facile reaction of vancomycin with various PEG linkers, at the V3 position, has been
selectively accomplished by using an excess of base in DMF. Using rPEG as a blocking group
for V3 provides crystalline derivatives that can be further PEGylated to give pure V3-X1
latentiated species (transport forms). V3 tetrameric species were also prepared in order to
increase the loading of drug on PEG. All PEG-vancomycin transport forms show significant
antibacterial activity that is on the same order of native vancomycin. Significant increases in
the AUC were observed for all PEG-vancomycin conjugates thus making them potential single
dose therapies.
Vancomycin (1) is a glycopeptide antibiotic that is the
drug of choice for the treatment of Gram-positive
infections caused by methicillin resistant Staphylococcus
aureus.1 It is also used in the treatment of bacterial
infections in patients allergic to â-lactam antibiotics.2
For the safe and effective use of this drug, quantitation
of its levels in patient’s blood is often required to
maintain therapeutic levels,3 and dosing is usually done
by infusion every 6 h for prolonged periods of time
(about 10 weeks) depending on the severity of the
infection being treated. Several groups have attempted
to enhance the performance of 1 by continuous
infusion,4-6 but it was concluded that no significant
enhancement of therapeutic efficacy was realized. How-
ever, it was observed6 that over a 10 day interval, the
cost of treatment per patient could be reduced by 30%
using the continuous infusion method.
To date, we are unaware of attempts to prepare long-
lived amine-based prodrugs of vancomycin (Vanco) that
could be utilized for limiting the number of infusions of
the drug required by the patient. Vancomycin prodrugs
would therefore serve to reduce associated nursing costs
and avoid the use of inserting an infusion pump, which
often can increase opportunistic infections at the im-
plant site. By definition, prodrugs are formed by modi-
fication of a chemical functionality on a drug that is
necessary for activity.7 In the case of Vanco it is not clear
if either of the V3 or X1 amino groups (Figure 1) is
necessary for activity since the degree of activity
observed can be altered in either direction by acylation
of one of these moieties8 and in part depends on the
structure of the acylating agent as well.9 Modification
of both positions does, however, appear to lower activ-
ity.9 The term latentiation was popularized in 1975 by
Sincula and Yalkowsky7 in order to incorporate the idea
of drug release from a conjugated species regardless of
the position of blockage. Irrespective of the type of labile
connection used, in the case of high molecular weight
PEG (>20 000) the conjugate could transport 1 through-
out the circulatory system for extended periods of time
before release. Thus, the objective of the present study
was to prepare PEG latentiated conjugates of 1, either
as prodrugs or transport forms, with the hope of
increasing the mean residence time of the drug in the
circulatory system, thereby limiting the number of
treatment infusions a patient would require. We were
encouraged in this endeavor by the recent report10 that
a prodrug approach to increase the circulating half-life
(t1/2) of the antibiotic gentamycin was successfully
carried out by conjugation of substituted Fmoc deriva-
tives with the available amino sugars on the molecule
and resulted in a continuous release of the native drug
while maintaining equivalent activity. The toxicity issue
of the sulfonated hydroxymethyl fluorene byproduct
formed during gentamycin release was not addressed,
but in the case of a PEG latentiated drug approach to
delivery, only the drug and nontoxic PEG species11
would be formed in vivo. Our aim then was to devise
unambiguous syntheses of labile PEG conjugates of 1,
establish in vivo efficacy, and, if warranted, determine
the rates of release (PK).
Ch em istr y
Previously conjugation of V3 and X1 to nonpolymeric
species was accomplished by reaction of 1 with chloro-
formates,12 acid anhydrides,12,13 active esters,9,14,15 and
also carboxylic acids in the presence of condensing
agents such as DCC14,15 or PyBOP.16 In all cases,
mixtures of both mono- and disubstituted acylated
products were obtained, but modulation of reaction
conditions were reported to favor reaction of one position
to a greater extent than the other. Isolation of pure
species could then be accomplished by preparative
HPLC. Thus, Adamczyk and co-workers14 using DCC
to condense carboxylic acids with 1 in DMSO were able
to isolate X1 monoacylated Vanco by preparative HPLC
from an enriched mixture of products. Unfortunately,
the direct condensation of PEG diacid17 with 1 using
the coupling reagents DCC, EDC, and DIPC in the
presence of DMAP produced only low yields of mixtures
that probably consisted of both mono- and diacylated
Vanco. Substitution of triethylamine (TEA) for DMAP
resulted in no reaction taking place.
* To whom correspondence should be addressed. Phone: (732) 980-
4924. Fax: (732) 885-2950. E-mail: richard.greenwald@enzon.com.
10.1021/jm030202g CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/03/2003

5022 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Greenwald et al.
F igu r e 1.
Sch em e 1
We have previously utilized 1,6-benzyl elimination
(1,6-BE) as a means of releasing amino-containing
bioactive substances from their PEG conjugates.18 Ap-
plication of this technology to Vanco would be flawed if
more than one major conjugated product was obtained,
for it is extremely difficult to separate mixtures of PEG
conjugates of different molecular weights. Thus, it
became of paramount importance to unearth reaction
conditions that would clearly differentiate the V3 and
X1 amino groups toward acylation not only for purifica-
tion factors, but because the use of a bifunctional PEG
linker could result in oligomerization.
A systematic study of reaction conditions based on
reported procedures9,14,15 was undertaken to optimize
reaction conditions for monofunctionalization of 1 using
the succinimidyl carbonate disubstituted PEG 40 000
linker 4 (ester trigger, Scheme 1), 15 and 16 (amino acid
spaced ester trigger, Scheme 2) and 24 (carbamate
trigger, Scheme 3). These acylation reagents were
prepared as shown in Schemes 1-4. Noteworthy in the
synthesis of PEG linker 24 (Scheme 3) was the use of
the novel carbonate reagent 20.19 This reagent was
designed to provide higher yields of the PEGylated
intermediate 22 than that provided by the reaction
between 12 and PEG-NCO (also see ref 19 for R ) H).
The coupling of PEG linker 4 to Vanco was investi-
gated in depth, and it was first determined that the use
of DMF as solvent was superior to other solvents
previously employed with Vanco such as DMSO (or even
mixtures of DMF-DMSO) for producing higher yields
of conjugates. Next, it was found that the presence, and
stoichiometry, of base was critical for maximization of
yields. When DIEA or DMAP were employed in the
reaction, low yields of disubstituted product resulted.
Some monosubstitution was observed when pyridine
was used as the base, but the most promising results
were obtained when TEA was employed. The results
gleaned from extensive investigation of using TEA in
this reaction resulted in the determination that at least
20 equiv of base per Vanco molecule were required to
optimize PEG conjugation. The use of less than 20 equiv
resulted in incomplete reaction. It was also determined
that NHS esters of PEG also reacted with Vanco in the
presence of 20 equiv of TEA to yield the corresponding
permanently bonded amide conjugate 3 in high yield
(Scheme 1), but surprisingly the amine specific PEG
linker, PEG-thiazolidine-2-thione (T-PEG),18 was almost
completely unreactive toward 1. Using the favorable
conditions established above for 1 with linkers 4, 15,
16, and 24, the resulting structures were determined
by NMR spectroscopy to be 5, 17, 18, and 25, respec-
tively. Thus selective monoacylation of the V3 position
by the various PEG 40 000 activated linkers had been
achieved. In the 13C NMR of the PEG derivatives
(pyridine-d5), the chemical shift (δ) of V3 [C-NH2]
changed from 53.9 ppm to about 52 ppm. Meanwhile δ
for X1 [C-NHCH3] remained intact at 62.5 ppm. The
purity of these products was >95% as determined by
HPLC and confirmed indirectly by GPC molecular
weight determination: no higher molecular weight
species were found to be present, which would be the
case if the X1 position had also reacted. With the

PEG Transport Forms of Vancomycin
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 5023
Sch em e 2
Sch em e 3
knowledge that V3 can be selectively acylated, we turned
our attention to the quandary of X1 position discrimina-
tion. It was apparent from some of our and others
unsuccessful approaches,9,15 that disubstituted product
was formed during coupling and therefore both the V3
and X1 positions are reactive. Since the V3 amino group
can be selectively acylated, it followed that this is the
more reactive of both positions under certain conditions,
and in theory could be protected while conditions for
the selected modification of the X1 position were deter-
mined. Following the removal of the V3 protecting group,
monoacylated X1 product theoretically could be isolated.
It was initially found that the PEG 40 000 conjugate 5
has a much greater organic solubility than 1, thus
making this approach quite feasible. We chose to use
an rPEG19,20 as the protecting group since conditions
for PEG conjugation were worked out and because the
facility of the synthesis permitted modification of the
trigger in order to differentiate release. Consequently,
the methyl-capped rPEG linker 26 (molecular weight
5000),20 reacted with 1 to produce the V3 acylated
derivative 27 as a crystalline solid that was, as in the
case of the higher MW analogue, substantially more
soluble in organic solvents than the parent. The utility
of using low MW PEG in this capacity is underscored
by the fact that attempts to prepare V3 conjugates with
small molecule blocking groups were not productive: the
resulting conjugated compounds exhibited very poor
solubility in both aqueous and organic solvents and
often gelled. Compound 27 in turn was dissolved in
DMF/DCM (1:1) as solvent and employing the acylation
catalyst, DMAP, further reaction with the disubstituted
PEG linker 4 (molecular weights 20 000 and 40 000)
gave the desired symmetrically disubstituted acylated
products, 28 and 29, in excellent yield (94%) as shown
in Scheme 4. It should be noted that both 28 and 29

5024 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Greenwald et al.
Sch em e 4
Sch em e 5
activated tetraacid derivative 33 (Scheme 5) with 12
followed by activation of the benzyl alcohol moiety with
NHS produced the tetrasubstituted 1,6-BE linker, 35.
Reaction of 1 equiv of 35 with 4 equiv of 1, as shown in
Scheme 5, gave the desired V3 substituted tetramer 36
with a loading of four vancomycins per PEG as deter-
mined by UV analysis.17,18 In this fashion, the % active
of Vanco was increased to about 10% (theoretical, 12%).
Similarly, a tetrameric Vanco V3 PEG carbamate
derivative, 43, was prepared as shown in Schemes 6.
Exp er im en ta l Section
Gen er a l P r oced u r es. All reactions were run under an
atmosphere of dry nitrogen or argon. Commercial reagents
were used without further purification. All PEG compounds
were dried under vacuum or by azeotropic distillation from
toluene prior to use. Poly(ethylene glycol) (PEG) linkers:
compounds 4, 26, and 30 were made according to the published
procedures.17-20 13C NMR spectra were obtained at 75.46 MHz
using a Varian Mercury 300 NMR spectrometer and deuter-
ated chloroform and pyridine as the solvents unless otherwise
specified. Chemical shifts (δ) are reported in parts per million
(ppm) downfield from tetramethylsilane (TMS).
HP LC Meth od s. The reaction mixtures and the purity of
intermediates and final products were monitored by a Beck-
man Coulter System Gold HPLC instrument. It employs a
ZOBAX 300SB C8 reversed phase column (150 × 4.6 mm) or
a Phenomenex J upiter 300A C18 reversed phase column (150
× 4.6 mm) with a multiwavelength UV detector, using a
gradient of 10-90% of acetonitrile in 0.05% trifluoroacetic acid
(TFA) at a flow rate of 1 mL/min.
Kin etic Stu d ies in Ra t P la sm a . PEG-vancomycin (either
20 mg of 40 kDa dimer conjugate, 10 mg of 20 kDa dimer
conjugate or 40 kDa tetramer conjugate) was dissolved in 1000
µL of a mixture of AcCN and MeOH (1:1, v/v). 100 µL of
solution was transferred to each vial (8 vials total, StepVial
System II). Solvent was removed under reduced pressure, and
100 µL of rat plasma was added to each vial. The vial was
sealed and vortexed for 1 min after which time the vial was
incubated at 37 °C for 0, 0.5, 1, 2, 4, 6, and 20 h, respectively.
To each vial was again added 400 µL of AcCN and MeOH (1:
1, v/v), and the vial was vortexed for 1 min. The mixture was
filtered through a 0.45 µm filter membrane, and 50 µL of the
filtrate was injected directly into the HPLC system.
are necessarily doubly latentiated transport systems.
Since the objective of modification of either the V3 and
X1 positions is to adjust the PK of latentiated Vanco,
then a double substitution should be a satisfactory
solution for that objective if it could be shown that Vanco
is regenerated from 28 and 29 under physiological
conditions. By using this concept, a difficult separation
of mPEG 5000 from an X1 PEG 40 000 conjugate can
be avoided. This view will be discussed in greater detail
in the Discussion section.
To decrease the volume of formulated drug at a given
concentration, resulting in decreased solution viscosity,
as previously discussed in the case of ara-C,21 it was
necessary to increase the loading of 1 onto the polymer.
Thus, tetrameric V3-PEG transport forms were prepared
using the known PEG scaffolding based on aspartic acid
as shown in Scheme 5.22 For instance conjugation of the
Exp er im en ta l P r oced u r es for th e Syn th eses of Va n -
com ycin -P EG Der iva tives. Com p ou n d 5. To a solution of

PEG Transport Forms of Vancomycin
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 5025
Sch em e 6
compound 1 (0.535 g, 0.369 mmol) and triethylamine (TEA,
1.026 mL, 7.36 mmol) in anhydrous dimethylformamide (DMF,
150 mL) was added PEG linker 4 (7.5 g, 0.184 mmol), and the
resulting mixture was stirred at room temperature for 12 h.
The solution was filtered through Celite and treated with ethyl
ether (300 mL). Filtration gave crude product which was
recrystallized from a mixture of DMF/TEA/2-propanol (IPA)
(45 mL/1 mL/180 mL) twice to give 5 (7.61 g, 0.175 mmol, 95%).
13C NMR (67.8 MHz, C5D5N) δ 174.44, 172.32, 170.65, 169.15,
168.74, 167.02, 159.02, 157.39, 155.37, 142.98, 142.09, 133.22,
130.37, 128.45, 127.32, 119.79, 118.28, 107.79, 102.76, 98.54,
90.25, 88.38, 78.93, 77.34, 76.41, 64.41, 62.58, 61.65, 60.71,
58.87, 56.97, 54.39, 53.35, 35.64, 25.29, 24.23, 23.48, 22.32,
18.46, 16.35, 12.08.
Com p ou n d 6. To a solution of 40 kDa PEG diacid (2, 22.0
g, 0.548 mmol), glycine tert-butyl ester hydrochloride (1.10 g,
6.58 mmol), and 4-(dimethylamino)pyridine (DMAP, 1.60 g,
13.16 mmol) in anhydrous methylene chloride (DCM, 200 mL)
cooled to 0 °C in an ice bath was added 1-(3-dimethylamino-
propyl)-3-ethylcarbodiimide hydrochloride (EDC, 1.26 g, 6.58
mmol). This mixture was allowed to warm to room tempera-
ture and stirred for 12 h, followed by partial removal of the
solvent under reduced pressure. The PEG derivative was
precipitated with addition of ethyl ether, filtered, and crystal-
lized from 2-propanol (IPA, 440 mL) to yield 6 (20.8 g, 0.515
mmol, 94%). 13C NMR (67.8 MHz, CDCl3) δ 169.92, 168.38,
70.36-69.78 (PEG), 40.85, 27.68.
Com p ou n d 7. To a solution of 2 (25.0 g, 0.623 mmol),
alanine tert-butyl ester hydrochloride (1.36 g, 7.48 mmol), and
DMAP (1.82 g, 14.95 mmol) in anhydrous DCM (300 mL)
cooled to 0 °C in an ice bath was added EDC (1.44 g, 7.48
mmol). This mixture was allowed to warm to room tempera-
ture and stirred for 12 h, followed by partial removal of the
solvent under reduced pressure. The PEG derivative was
precipitated by the addition of ethyl ether, filtered, and
crystallized from IPA (500 mL) to yield 7 (20.8 g, 0.515 mmol,
94%). 13C NMR (67.8 MHz, CDCl3) δ 171.19, 168.78, 71.05-
69.32 (PEG), 47.62, 27.64, 18.10.
at room temperature for 2 h, followed by partial removal of
the solvent under reduced pressure. The PEG derivative was
precipitated by the addition of ethyl ether, filtered, and washed
with ethyl ether to yield 8 (18.2 g, 0.451 mmol, 91%). 13C NMR
(67.8 MHz, CDCl3) δ 170.07, 170.00, 72.91-67.29 (PEG), 39.99.
Com p ou n d 9. A solution of 7 (23.0 g, 0.570 mmol) in DCM
(250 mL) and TFA (150 mL) was stirred at room temperature
for 2 h, followed by partial removal of the solvent under
reduced pressure. The PEG derivative was precipitated by the
addition of ethyl ether, filtered, and washed with ethyl ether
to yield 9 (22.6 g, 0.559 mmol, 98%). 13C NMR (67.8 MHz,
CDCl3) δ 172.67, 168.98, 71.02-69.83 (PEG), 46.89, 17.93.
Com p ou n d 10. To a solution of 8 (17.0 g, 0.423 mmol),
2-mercaptothiazoline (2-MT, 0.20 g, 1.69 mmol), and DMAP
(0.21 g, 1.69 mmol) in anhydrous DCM (150 mL) cooled to 0
°C was added EDC (0.32 g, 1.69 mmol). The reaction mixture
was allowed to warm to room temperature and stirred for
12 h, followed by partial removal of the solvent under reduced
pressure. The PEG derivative was precipitated by the addition
of ethyl ether, filtered, and crystallized from IPA (340 mL) to
yield 10 (16.5 g, 0.410 mmol, 97%). 13C NMR (67.8 MHz,
CDCl3) δ 200.96, 170.22, 170.01, 72.13-69.72 (PEG), 55.35,
45.58, 28.67.
Com p ou n d 11. To a solution of 9 (22.5 g, 0.558 mmol),
2-MT (0.266 g, 2.24 mmol), and DMAP (0.27 g, 2.24 mmol) in
anhydrous DCM (250 mL) cooled to 0 °C was added EDC (0.43
g, 2.24 mmol). The reaction mixture was allowed to warm to
room temperature and stirred for 12 h, followed by partial
removal of the solvent under reduced pressure. The PEG
derivative was precipitated with the addition of ethyl ether,
filtered, and crystallized from IPA (450 mL) to yield 11 (21.3
g, 0.525 mmol, 94%). 13C NMR (67.8 MHz, CDCl3) δ 200.41,
174.12, 168.92, 72.93-69.34 (PEG), 56.00, 48.03, 28.45, 17.85.
Com p ou n d 13. A solution of 10 (16.0 g, 0.40 mmol), 3,5-
dimethyl-4-hydroxybenzyl alcohol (0.78 g, 5.14 mmol), and
DMAP (0.63 g, 5.14 mmol) in anhydrous DCM (150 mL) was
refluxed for 18 h, followed by partial removal of the solvent
under reduced pressure. The PEG derivative was precipitated
with addition of ethyl ether, filtered, and crystallized from IPA
to yield 13 (14.6 g, 0.36 mmol, 90%). 13C NMR (67.8 MHz,
Com p ou n d 8. A solution of 6 (20.0 g, 0.496 mmol) in DCM
(200 mL) and trifluoroacetic acid (TFA, 100 mL) was stirred

5026 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Greenwald et al.
CDCl3) δ 170.53, 167.47, 146.49, 138.80, 129.52, 126.61, 72.14-
69.54 (PEG), 63.68, 39.96, 15.89.
(4 mL) cooled to 0 °C was added pyridine (0.125 g, 1.60 mmol).
The reaction mixture was allowed to warm to room temper-
ature and stirred for 12 h, followed by partial removal of the
solvent under reduced pressure. The PEG linker was precipi-
tated by addition of ethyl ether, filtered, and crystallized from
DCM/ethyl ether to yield 24 (3.45 g, 0.85 mmol, 86%). 13C NMR
(67.8 MHz, CDCl3) δ 169.44, 151.34, 128.14, 126.65, 123.56,
72.38-68.90 (PEG), 41.52, 25.08, 15.79.
Com p ou n d 25. Prepared in 86% yield by reacting 1 with
24 as described for 5. 13C NMR (67.8 MHz, C5D5N) δ 174.44,
172.41, 170.65, 168.74, 166.18, 158.76, 157.45, 154.24, 152.74,
152.10, 148.09, 137.54, 128.77, 127.30, 90.29, 88.42, 78.84,
77.37, 76.43, 63.50, 62.66, 61.65, 60.74, 58.87, 56.99, 55.16,
54.20, 53.35, 52.81, 51.39, 35.64, 26.12, 25.26, 23.51, 22.30,
18.47, 17.43, 14.00.
Com p ou n d 14. A solution of 11 (12.4 g, 0.30 mmol), 12
(0.78 g, 5.14 mmol), and DMAP (0.63 g, 5.14 mmol) in
anhydrous DCM (150 mL) was refluxed for 18 h, followed by
partial removal of the solvent under reduced pressure. The
PEG derivative was precipitated with addition of ethyl ether,
filtered, and crystallized from IPA to yield 14 (11.8 g, 0.29
mmol, 95%). 13C NMR (67.8 MHz, CDCl3) δ 170.00, 169.38,
146.36, 138.71, 129.45, 126.62, 72.93-69.06 (PEG), 63.80,
47.30, 17.73, 16.04.
Com p ou n d 15. To a solution of 13 (3.8 g, 0.094 mmol) and
disuccinimidyl carbonate (DSC, 0.19 g, 0.75 mmol) in anhy-
drous DCM (40 mL) and DMF (4 mL) cooled to 0 °C was added
pyridine (0.059 g, 0.75 mmol). The reaction mixture was
allowed to warm to room temperature and stirred for 12 h,
followed by partial removal of the solvent under reduced
pressure. The PEG linker was precipitated by the addition of
ethyl ether, filtered, and crystallized from DCM/ethyl ether
to yield 15 (3.3 g, 0.082 mmol, 87%). 13C NMR (67.8 MHz,
CDCl3) δ 170.52, 168.15, 167.24, 151.06, 148.02, 130.55, 130.42,
128.47, 71.83-69.64 (PEG), 40.12, 25.18, 16.05.
Com p ou n d 16. To a solution of 14 (9.0 g, 0.222 mmol) and
DSC (1.14 g, 4.44 mmol) in anhydrous DCM (90 mL) and DMF
(9 mL) cooled to 0 °C was added pyridine (0.351 g, 4.44 mmol).
The reaction mixture was allowed to warm to room temper-
ature and stirred for 12 h, followed by partial removal of the
solvent under reduced pressure. The PEG linker was precipi-
tated by addition of ethyl ether, filtered, and crystallized from
DCM/ethyl ether to yield 16 (6.5 g, 0.158 mmol, 71%). 13C NMR
(67.8 MHz, CDCl3) δ 170.05, 169.64, 168.20, 151.18, 148.07,
130.67, 130.51, 128.65, 71.98-69.93 (PEG), 47.48, 25.30, 17.82,
16.21.
Com p ou n d 17. Prepared in 88% yield by reacting 1 with
15 as described for 5. 13C NMR (67.8 MHz, C5D5N) δ 174.68,
171.17, 168.62, 167.14, 159.00, 157.50, 155.50, 142.00, 133.50,
130.51, 128.38, 127.24, 78.93, 65.22, 63.76, 62.49, 54.39, 52.20,
44.61, 43.29, 35.52, 25.23, 24.20, 23.46, 22.30, 18.47, 16.37.
Com p ou n d 18. Prepared in 83% yield by reacting 1 with
16 as described for 5. 13C NMR (67.8 MHz, C5D5N) δ 174.80,
171.46, 170.77, 144.20, 130.84, 128.75, 127.59, 119.00, 79.10,
78.38, 64.95, 62.92, 61.99, 54.69, 52.31, 51.38, 48.89, 35.85,
25.57, 24.52, 23.78, 22.65, 18.75, 16.79, 15.27.
Com p ou n d 27. Prepared in 91% yield by reacting 1 with
26 as described for 5.
Com p ou n d 28. A solution of 4 (mw 20664, 4.44 g, 0.215
mmol), 27 (mw 5332, 2.73 g, 0.410 mmol), and DMAP (0.251
g, 2.05 mmol) in a mixture of anhydrous DCM:DMF (50 mL:
50 mL) was stirred at room temperature for 12 h. The PEG
derivative was precipitated by the addition of ethyl ether (300
mL) and filtered, and the residue was recrystallized from a
mixture of ethanol:DMF:TEA (150 mL:150 mL:3.0 mL) twice
to give 28 (6.82 g, 0.202 mmol, 94%). 13C NMR (67.8 MHz,
C5D5N) δ 174.68, 171.94, 170.51, 168.73, 168.62, 162.35,
155.42, 141.02, 133.16, 132.06, 130.69, 130.37, 130.08, 128.45,
127.29, 78.08, 72.5-69.0 (PEG), 65.24, 63.77, 58.88, 54.48,
35.89, 31.02, 26.21, 25.28, 24.18, 23.46, 22.33, 18.29, 16.35.
Com p ou n d 29. A solution of 4 (mw 40,664, 3.84 g, 0.0945
mmol), 27 (mw 5332, 1.20 g, 0.180 mmol), and DMAP (0.251
g, 2.05 mmol) in a mixture of anhydrous DCM:DMF (60 mL:
60 mL) was stirred at room temperature for 12 h. The PEG
derivative was precipitated by the addition of ethyl ether (300
mL), filtered, and the residue was crystallized from a mixture
of ethanol:DMF:TEA (150 mL:150 mL:1.2 mL) to give 29 (6.82
g, 0.202 mmol, 94%).
Com p ou n d 31. To a solution of 30 (30.0 g, 0.744 mmol),
glycine methyl ester hydrochloride (0.75 g, 5.95 mmol), and
DMAP (1.8 g, 14.75 mmol) in DCM (300 mL) cooled to 0 °C
was added EDC (1.71 g, 8.92 mmol), and the mixture was
allowed to warm to room temperature and stirred overnight,
followed by partial removal of the solvent in vacuo. The product
was precipitated with ethyl ether, collected by filtration, and
crystallized from IPA (600 mL) to give 31 (29.2 g, 97%). 13C
NMR (67.8 MHz, C5D5N) δ 170.37, 170.18, 169.82, 169.57,
169.31, 70.61-69.93 (PEG), 51.78, 48.87, 40.84, 40.80, 36.77.
Com p ou n d 32. A solution of 31 (29 g, 0.71 mmol) and
lithium hydroxide hydrate (0.24 g, 5.71 mmol) in water (200
mL) was stirred overnight at room temperature and the pH
adjusted to 2.5 with 0.1 N HCl. This solution was extracted
with DCM and the organic layer dried (anhydrous sodium
sulfate) and filtered, followed by partial removal of the solvent
under reduced pressure. The product was precipitated using
ethyl ether, collected by filtration, and crystallized from IPA
(640 mL) to give 32 (27.4 g, 94%). 13C NMR (67.8 MHz, C5D5N)
δ 170.44, 170.32, 170.11, 169.86, 72.16-69.78 (PEG), 61.26,
49.22, 41.09, 37.30.
Com p ou n d 22. To a solution of 19 (0.73 g, 4.89 mmol) and
triphosgene (0.21 g, 0.70 mmol) in 100 mL of dry DCM cooled
to 15 °C was added diisopropylamine (DIEA, 1.02 mL) drop-
wise over a period of 5 min. The reaction mixture was then
allowed to warm to room temperature and stirred for 1 h to
give compound 20,19 which was used without further purifica-
tion. The reagents, 40 kDa PEG diamine hydrochloride (7.0
g, 0.175 mmol) and DIEA (0.061 mL) were added to the above
solution, and the resulting mixture was stirred for 18 h at room
temperature followed by partial removal of the solvent under
reduced pressure. The PEG derivative was precipitated by the
addition of ethyl ether, filtered, and crystallized from IPA to
yield 22 (5.6 g, 0.138 mmol, 79%). 13C NMR (67.8 MHz, C5D5N)
δ 190.69, 162.10, 130.31, 128.21, 124.05, 123.57, 72.11-69.49
(PEG), 39.92, 15.72.
Com p ou n d 23. To a solution of 22 (5.5 g, 0.14 mmol) in
methanol (70 mL) cooled to 15 °C was added sodium borohy-
dride (0.018 g, 0.45 mmol). The reaction mixture was allowed
to warm to room temperature over a period of 2 h, followed by
adjusting the pH to 6.5 with 1 N HCl. The solvent was removed
under reduced pressure and the residue taken up in water.
The pH was lowered to 2.0 with 0.5 N HCl and the product
extracted from the water with DCM. The organic layer was
dried over anhydrous sodium sulfate and filtered, followed by
partial removal of the solvent under reduced pressure. The
product was precipitated by addition of ethyl ether, filtered,
and washed with ethyl ether to yield 23 (4.7 g, 0.12 mmol,
85%). 13C NMR (67.8 MHz, C5D5N) δ 153.52, 130.12, 124.12,
121.33, 73.67-69.51 (PEG), 65.47, 39.91, 15.86.
Com p ou n d 33. To a solution of 32 (5.0 g, 0.123 mmol),
2-MT (0.18 g, 1.51 mmol), and DMAP (0.36 g, 2.96 mmol) in
anhydrous DCM (50 mL) cooled to 0 °C was added EDC (0.28
g, 1.48 mmol) and stirred overnight, followed by partial
removal of the solvent under reduced pressure. The product
was precipitated with ethyl ether, collected by filtration, and
crystallized from IPA (100 mL) to yield 33 (4.6 g, 92%). 13C
NMR (67.8 MHz, C5D5N) δ 200.74, 200.71, 170.29, 170.17,
169.85, 169.64, 72.09-68.39 (PEG), 55.39, 48.92, 46.32, 36.67,
28.72.
Com p ou n d 34. A solution of 33 (4.5 g, 0.11 mmol), 3,5-
dimethyl-4-hydroxybenzyl alcohol (0.35 g, 2.87 mmol), and
DMAP (0.53 g, 2.87 mmol) in anhydrous DCM (50 mL) was
refluxed for 18 h, followed by partial removal of the solvent
under reduced pressure. The PEG derivative was precipitated
by the addition of ethyl ether, filtered, and crystallized from
Com p ou n d 24. To a solution of 23 (4.0 g, 0.99 mmol) and
DSC (0.41 g, 1.60 mmol) in anhydrous DCM (40 mL) and DMF

PEG Transport Forms of Vancomycin
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 5027
IPA (90 mL) to yield 34 (4.1 g, 91%). 13C NMR (67.8 MHz,
CDCl3) δ 170.69, 170.38, 169.96, 167.36, 167.06, 148.39, 138.72,
129.46, 126.56, 72.15-69.17 (PEG), 63.77, 49.01, 40.76, 36.83,
16.02.
130.04, 128.53, 72.21-68.18(PEG), 49.40, 40.96, 39.18, 37.19,
25.29, 16.11.
Com p ou n d 43. This was prepared in 79% yield by reacting
1 with 42 as described for 36. HPLC purity 98% (method
described in the General Procedures).
Com p ou n d 35. To a solution of 34 (4.0 g, 0.097 mmol) and
DSC (0.80 g, 3.11 mmol) in anhydrous DCM (40 mL) and DMF
(4 mL) cooled to 0 °C was added pyridine (0.25 g, 3.11 mmol).
The reaction mixture was allowed to warm to room temper-
ature and stirred for 12 h followed by partial removal of the
solvent under reduced pressure. The PEG linker was precipi-
tated by the addition of ethyl ether, filtered, and crystallized
from DCM/ethyl ether to yield 35 (2.8 g, 70%). 13C NMR (67.8
MHz, CDCl3) δ 170.73, 170.43, 169.97, 168.17, 167.18, 166.86,
151.03, 147.96, 130.55, 130.38, 128.42, 71.78-69.78 (PEG),
49.05, 40.78, 36.92, 25.15, 15.98.
Meth od s a n d Resu lts
LD90-100 Exa m in a tion of P EG-Va n com ycin Con -
ju ga tes 5, 17, 18, 25, 28, 29, 36, a n d 43 in ICR Sw iss
Mice. Studies examining the in vivo anti-infective
potential of PEG-vancomycin conjugates in ICR Swiss
mice were performed. In the initial study, PEG-
vancomycin conjugates 5, 17, 18, 25, 28, 29, 36, and 43
were administered at 100 mg/kg vancomycin equiva-
lents in two equal split doses given 1 and 2.5 h following
a LD90-100 inoculation of Staphylococcus aureus (Smith)
to ICR Swiss mice. As controls, S. aureus (Smith)-
challenged mice were administered either saline or 100
mg/kg vancomycin given in equal split doses. Ten mice
were used per compound tested. Mice were observed for
1 week for survival. Significant antimicrobial activity
was indicated by greater than 50% survival of the mice.
(Table 4).
A second study was also performed to examine the
sustained anti-infective protection of PEG-vancomycin
conjugate 18. ICR Swiss mice were administered with
either 100 mg/kg vancomycin equivalents of 18, 100 mg/
kg vancomycin, or saline in two equally split doses 1.5
h apart. Ten mice per treatment group were used. Mice
were challenged with an LD90-100 inoculum of S. aureus
(Smith) 8 h after initial administration of test com-
pounds. Mice were observed for survival as before.
The studies showed that PEG-vancomycin conju-
gates 5, 17, 18, 25, 28, 29, 36, and 43 when adminis-
tered at 100 mg/kg vancomycin equivalents showed
significant activity against S. aureus-challenged mice.
Likewise, the PEG-vancomycin conjugate 18 showed
sustained 8 h antimicrobial protection against S. aureus
challenge in ICR Swiss mice and significant antimicro-
bial activity with 80% survival of the mice. Native
vancomycin treatment and saline control showed no
protection with 0% survival. All the data are sum-
marized in Table 4.
Com p ou n d 36. To a solution of 1 (0.353 g, 0.238 mmol)
and TEA (0.662 mL, 4.75 mmol) in anhydrous DMF (50 mL)
was added 35 (2.2 g, 0.0528 mmol), and the resulting mixture
was stirred at room temperature for 12 h. The reaction mixture
was filtered through Celite and the product precipitated with
ethyl ether (300 mL). Filtration gave crude product which was
recrystallized twice from a mixture of DMF/ethanol (1:1) to
give 36 (2.0 g, 90%). HPLC purity, 98% (method described in
the General Procedures).
Com p ou n d 38. To a solution of 37 (4.0 g, 16.1 mmol) and
triphosgene (1.92 g, 6.46 mmol) in anhydrous DCM (50 mL)
cooled to 15 °C was added DIEA (7.6 mL, 43.6 mmol) dropwise
over a period of 5 min. This mixture was allowed to warm to
room temperature over a period of 1 h, followed by the addition
of 20 (1.97 g, 16.13 mmol) and DMAP (2.4 g, 16.1 mmol), and
then stirred at room-temperature overnight. The mixture was
washed with 0.1 N HCl solution, the organic layer dried
(anhydrous sodium sulfate) and filtered, and the solvent
removed from the filtrate under reduced pressure to give crude
38. 13C NMR (67.8 MHz, CDCl3) δ 191.18, 155.64, 152.99,
152.82, 133.42, 132.06, 129.76, 78.99, 41.00, 40.17, 28.21,
16.12.
Com p ou n d 39. To a solution of 38 (1.0 g, 2.36 mmol) in
methanol (30 mL) cooled to 15 °C was added sodium borohy-
dride (0.1 g, 2.63 mmol), and the reaction mixture was stirred
at room temperature for 1 h, followed by acidification with 0.1
N HCl solution. The solvent was removed from the filtrate in
vacuo, and the residue was taken up in water (20 mL) and
extracted with DCM. The organic layer was dried (anhydrous
sodium sulfate) and filtered and the solvent removed under
reduced pressure. The crude product was purified by silica gel
chromatography to yield 39 (0.9 g, 90%). 13C NMR (67.8 MHz,
CDCl3) δ 155.65, 153.95, 146.81, 138.01, 130.33, 126.56, 78.86,
63.89, 40.69, 39.96, 28.09, 15.92.
P h a r m a cok in etics of Va n com ycin (1) a n d P EG-
Va n com ycin Con ju ga tes 17, 18, 25, 28, 29, 36, a n d
43 in Ra ts. This study was performed to determine the
circulatory pharmacokinetics of PEG-vancomycin con-
jugates (17, 18, 25, 28, and 29) in rats. PEG-vanco-
mycin conjugates containing a 50 mg/kg vancomycin
equivalent dose and 50 mg/kg native vancomycin (1)
were injected as a single intravenous dose into the tail
vein of conscious rats at a rate of 0.5 mL/min [∼2.5 min].
Blood samples were obtained 72 h prior to treatment
and at 0.08, 0.5, 1, 2, 4, 8, 24, 48, 72, and 96 h. A 250
µL bleed was done on rats sedated with 30% O2/70%
CO2 via the retro-orbital plexus into EDTA containing
vials. Blood was immediately processed for plasma and
frozen on dry ice. Plasma samples were stored at -80
°C until analyzed. The plasma samples were analyzed
for free vancomycin by a fluorescence polarization
immunoassay using an Abbott Laboratories TDx ana-
lyzer within 2 h of thawing. A single compartment
intravenous, first-order elimination model using Win-
Nonlin software was used to determine the plasma
pharmacokinetic parameters for conjugates 1, 28, and
Com p ou n d 41. To a solution of 39 (0.9 g, 2.11 mmol) in
DCM (57 mL) was added trifluoroacetic acid (3 mL), and the
reaction mixture was stirred for 1.5 h at room temperature
followed by removal of the solvent in vacuo. The residue was
dissolved in anhydrous DCM (70 mL), and to this solution was
added 30 (7.4 g, 0.18 mmol) and DMAP (1.26 g, 10.3 mmol).
The mixture was cooled to 0 °C followed by the addition of
EDC (0.56 g, 2.93 mmol). The solution was allowed to warm
to room temperature and stirred overnight followed by partial
removal of the solvent under reduced pressure. The product
was precipitated with ethyl ether, collected by filtration, and
crystallized from DCM/ethyl ether to yield 41 (6.45 g, 85%).
13C NMR (67.8 MHz, CDCl3) δ 170.34, 170.23, 169.82, 153.96,
147.06, 146.91, 138.39, 138.02, 130.61, 130.55, 126.79, 72.30-
69.29(PEG), 64.35, 64.26, 49.17, 39.19, 39.05, 16.16.
Com p ou n d 42. To a solution of 41 (6.4 g, 0.154 mmol) and
DSC (1.26 g, 4.92 mmol) in a mixture of anhydrous DCM (120
mL) and DMF (12 mL), cooled to 0 °C, was added pyridine
(0.39 g, 4.92 mmol), and the mixture was allowed to warm to
room temperature and stirred overnight. The solvent was
partially removed under reduced pressure and the product
precipitated with ethyl ether, filtered, and crystallized from
DCM/ethyl ether to give 42 (5.8 g, 89%). 13C NMR (67.8 MHz,
CDCl3) δ 170.26, 169.91, 168.25, 153.59, 151.15, 148.51, 131.31,

5028 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Greenwald et al.
Ta ble 1. Properties of PEG-Vancomycin Conjugates (V3
Dimers)
stability in
% of
solubilityb saline (24 h), t1/2 (rp) t1/2 (hp)
c
compd mw Vanco×bb (mg/mL)
%
(h)
(h)
2
5
17
18
25
42936
43332
42669
43457
43275
6.75
6.60
6.56
6.10
5.15
128
131
145
122
141
,1
<3
<2
<2
<1
.24
2
19
16
>24
- -
6.6
27
>24
.24
×bb
b
Percent of vancomycin by weight. Solubility was measured
in saline at 25 °C. c Percent of hydrolysis after 24 h at 25 °C.
Ta ble 2. Properties of PEG-Vancomycin Conjugates
(X1-Substituted V3 Dimers)
% of solubilityb stability in
Vanco (mg/mL) saline (24 h)c
t1/2 (rp) t1/2 (hp)
compd
mw
(h)
(h)
28
29
33764 8.80
53764 5.39
180
128
<1
<1
4.6
5.0
32
22
F igu r e 2. Vancomycin plasma concentration-time curve
(dimers). Rats were administered 50 mg/kg of compound
intravenously and bled 250 µL at 5 min, 1, 2, 4, 8, 24, 38, 72,
and 96 h via retro-orbital plexus. Plasma was analyzed for
vancomycin concentration by a fluorescent polarization im-
munoassay.
a
b
Percent of vancomycin by weight. solubility was measured
in saline at 25 °C. c Percent of hydrolysis after 24 h at 25 °C.
29 and a two-compartment intravenous, first-order
elimination model for conjugates 17, 18, 25, 36, and 43.
The plasma concentration-time curves showed coef-
ficients of determination (r2) of greater than 0.96 for all
PEG-vancomycin conjugates and 1.00 for vancomycin
itself.
Ta ble 3. Properties of PEG-Vancomycin Conjugate (V3
Tetramer)
% of solubilityb stability in
Vanco (mg/mL) saline (24 h)c
t1/2 (rp) t1/2 (hp)
compd
mw
(h)
(h)
In this study, vancomycin showed a circulatory half-
life (t1/2) of 0.34 h with a Cmax of 162 µg/mL and an area
under the curve (AUC) of 78.8 h‚µg/mL. Vancomycin had
a clearance (CL) of 642 mL/h/kg and a volume of
distribution at a steady state (Vss) of 309 mL/kg in rats.
Overall the pharmacokinetic estimates of the PEG-
vancomycin conjugates uniformly showed a longer sus-
tained, although lower concentration, of circulating free
vancomycin in rat plasma. All of the PEG-vancomycin
conjugates showed a 12 to 64-fold longer t1/2 (4.08 h to
21.9 h). All of the conjugates achieved Cmax values
that were 16% to 66% (26 µg/mL to 107 µg/mL) of
those observed for unmodified vancomycin. The AUC
for the PEG-vancomycin conjugates were 3.6-27.7 fold
greater (282.2 h‚µg/mL to 2184 h‚µg/mL) than that
observed for vancomycin and had concomitantly slower
clearance rates (71 mL/h/kg to 293.9 mL/h/kg) that were
11% to 45% of vancomycin’s. The Vss of the PEG-
vancomycin conjugates were 3-9-folds greater (938 mL/
kg - 2723 mL/kg) than that observed for vancomycin.
The PK profile is summarized in the following Table 5.
36
43
47011
47483 10.4
9.06
120
141
<2%
<1%
20
36
>24
.24
a
b
Percent of vancomycin by weight. Solubility was measured
in saline at 25 °C. c Percent of hydrolysis after 24 h at 25 °C.
Ta ble 4. Compound Description and Summary of LD90-100
Results
posta challenge
% survival
prechallenge
% survival
compound
saline
1
10
100
100
100
100
90
100
80
100
100
0
0
5
- -
- -
80b
- -
- -
- -
- -
- -
17
18
25
28
29
36
43
a
Compound given 1h following S. aureus (Smith) challenge,
b
survival after 1 week. Compound given 8 h prior to S. aureus
(Smith) challenge, survival after 1 week (single experiment).
since its CL is the highest observed in the study (Table
5, Figure 2). This may be the result of more rapid renal
and endoreticular clearance which is known to occur as
the MW of PEG changes.23
Discu ssion
After completion of the synthesis of the various
latentiated disubstitued 40 000 mw PEG V3 derivatives
of 1 as described above, the physical properties of these
derivatives were determined and are listed in Table 1.
Very little hydrolysis occurs for all derivatives in saline
(pH 7.0) thus enabling stable formulations to be pre-
pared prior to use. However, the rate of decomposition
(t1/2) in rat plasma for the PEG transport forms varies
significantly from 1 to >24 h: a similar trend is
observed in the case of human plasma (Tables 1-3).
But, as shown in Table 4, efficacy for all the PEG
conjugates is virtually equipotent to that of Vanco itself.
It can be seen that 5-6 wt % of 1 is present in the
PEG 40 000 conjugates, as shown in Table 2. Substitu-
tion of PEG 20 000 which provides higher loading (8.8%,
compound 28, Table 2) results in a very poor PK profile
Compound 18, with a 4-fold greater AUC than 1, was
chosen at random for a prechallange experiment. Par-
ticularly noteworthy is that 18, with an in vitro t1/2 of
16 h (PK t1/2 4.1 h) shows significant activity (80%
survival) when given 8 h prior to bacterial challenge (S.
aureus), while native 1 demonstrates 0% survival.
A comparison of the similar conjugates 17 (gly ester
trigger) and 18 (ala ester trigger), which have very
similar in vitro properties, shows a dissimilar in vivo
plasma t1/2. This may be due to an enzyme specific
cleavage of the alanine derivative. Conjugate 25 (PEG
40 kDa, carbamate trigger,V3 substitution) decomposes
due to the base-catalyzed hydrolysis of a phenolic
carbamate18 and results in a t1/2 of 9.1 h and gives rise

PEG Transport Forms of Vancomycin
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23 5029
Ta ble 5. Pharmacokinetics of Intravenous Bolus Administration of 50 mg/kg Vancomycin Equivalents of PEG-Vancomycin
Conjugates in Rats
compound
Cmax (µg/mL)
plasma t1/ 2 (hr)
CL (mL/h/kg)
Vss (mL/kg)
AUC (h‚µg/mL)
1
162.0 ( 9.0
25.5 ( 2.1
47.9 ( 4.3
54.6 ( 2.5
53.7 ( 5.4
24.8 ( 7.7
29.3 ( 5.1
107.1 ( 16.5
0.34 ( 0.04
10.48 ( 0.58
4.09 ( 0.25
9.12 ( 0.98
2.23 ( 0.38
21.19 ( 6.90
12.37 ( 2.18
14.2 ( 1.40
642.0 ( 83.7
131.1 ( 18.2
178.3 ( 13.4
70.0 ( 5.4
309.3 ( 17.2
2723.3 ( 97.3
2640.8 ( 23.9
1648.3 ( 81.5
938.0 ( 94.3
2128.6 ( 560.1
3598 ( 462
78.8 ( 10.8
386.4 ( 51.4
282.2 ( 28.5
716.9 ( 56.8
171.0 ( 15.2
708.4 ( 61.0
513 ( 40
17
18
25
28
29
36
43
293.9 ( 26.1
70.9 ( 6.3
97.8 ( 7.3
23.0 ( 2.1
1341 ( 52.5
2184 ( 189
to the largest AUC found for dimeric PEG-vanco
latentiated derivatives in this study (Table 5, Figure 2).
Pharmacokinetic (PK) values for most of the conjugates
were determined as described in the Methods and
Results section and are shown in Table 5. Examination
of the PK data indicate an impressive increase in
circulating half-life and a much larger AUC for all PEG
conjugates.
Conjugation (blocking) of theV3 position with 26, gave
latentiated 27 (R-alkoxyester trigger), which was further
functionalized at the X1 position with a PEG 40 000 to
yield the Vanco transport form, 29 (double ester trigger,
X1-protected with V3 substitution). This conjugate has
a very similar AUC as that found for 25 and an
extremely long plasma t1/2 (21.2 h, Table 5, Figure 2).
The conjugate 29 while providing an excellent PK profile
(Table 5) carries only about 5% of 1 because of the
increase in MW. Note that an approximate t1/2 for X1
PEG conjugates can be arrived at by comparing the V3
substituted PEG 40 000 compound 5, t1/2 )2 h, with 29
(V3 protection and X1 substitution), t1/2 ) 5 h, (Tables 1
and 2). Fortunately, since 29 had a longer t1/2, then it
follows that this will also be the rate-determining step
in the breakdown of the X1 derivative itself. Thus, there
is no need to remove the V3 blocking PEG group in order
to investigate an X1 latentiated Vanco derivative-
especially in light of the fact that the lower MW PEG
(5,000) can hydrolyze either at the same or more rapid
rate, but never at a slower rate.17
F igu r e 3. Vancomycin plasma concentration-time curve
(tetramers). Rats were administered 50 mg/kg of compound
intravenously and bled 250 µL at 5 min, 1, 2, 4, 8, 24, 48, 72,
and 96 h via retro-orbital plexus. Plasma was analyzed for
vancomycin concentration by fluorescent polarization immu-
noassay.
Con clu sion s
Exclusive V3 conjugation of PEG with Vanco can be
done in DMF as solvent with a large excess of TEA.
Although the same selectivity could not be demonstrated
for the X1 position, it was found that low MW PEG could
be used effectively as a blocking group at the V3 position,
which then allowed PEG conjugation of the X1 position
to take place. Furthermore, doubly latentiated PEG-
Vanco derivatives (V3-X1 substitution) had a greater t1/2
than the V3 conjugates, thus demonstrating that hy-
drolysis of X1 in the V3-X1 transport form is in fact the
rate controlling step in the plasma kinetic study.
Therefore, the PK profile of V3-X1 is essentially equiva-
lent to X1 substituted derivatives alone. The more highly
loaded V3 tetramer is preferable for further development
since the synthetic routes are more facile than the
preparation of the double latentiated forms. The current
protocols for Vanco administration gives a rapid peak
concentration (spike) that is rapidly cleared from the
bloodstream and thus results in the need for mutidosing
regimens. PEG-Vanco transport forms may provide a
novel and useful form of Vanco delivery, because slow
release of Vanco from these conjugates can be used in
place of continuous infusion type therapies. This ap-
proach now offers a drug delivery system that is
convenient, efficacious, less painful to the patient and
at dramatically reduced costs. Predosing experiments
have clearly demonstrated that while native 1 has no
effect under these conditions, latentiated 1 provides
short range protection from future bacterial exposure.
Increased loading adjustments of the PEG 40 000
conjugates were therefore done by preparing tetramers
whose water solubilities are still adequate for formula-
tion (Table 3). For a decreased dosing schedule to be
perfected for the various conjugates, the variables to be
considered would include a combination of high loading,
low Cmax (no spiking), and long t1/2. Generally, it is far
simpler to directly synthesize the V3 substituted com-
pounds; therefore further conversions to the more highly
loaded tetrameric forms were based on this consider-
ation. From Table 5 it is clear that the two dimeric
compounds with the greatest AUC, longest t1/2, and
lowest CL were 17 and 25. Also from inspection of
Figure 2 detectable amounts of 1 could still be observed
to be greater than 2 µg/mL for both conjugates after 100
h. On the basis of these factors the tetrameric latenti-
ated conjugate, 36 analogous to 17 was synthesized as
shown in Schemes 5. The tetrameric compound 36,
utilizing a glycine ester trigger with a loading of 9%
Vanco, had an AUC of 513, a reasonable t1/2 of 12.4 h,
and a relatively low Cmax of 29.3 µg/mL. A second
tetramer, 43 (analogous to 25), with a 10% loading of
1, as expected (43) had a large AUC of >2000 (Table 3,
Figure 3).

5030 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 23
Greenwald et al.
(13) Perkins, H. R. Specificity of combination between mucopeptide
precursors and vancomycin or ristocetin. Biochem. J . 1969, 111,
195-205.
(14) Adamczyk, M.; Grote, J .; Moore, J . A.; Rege, S. D.; Yu, Z.
Structure-binding relationships for the interaction between a
vancomycin monoclonal antibody fab fragment and a library of
vancomycin analogues and tracers. Bioconjugate Chem. 1999,
10, 176-185.
(15) Ghosh, M.; Miller, M. J . Synthesis and in vitro antibacterial
activity of spermidine-based amixed catechol- and hydroxamate-
containing siderophore-vancomycin conjugates. Bioorg. Med.
Chem. 1996, 4, 43-48.
(16) Staroske, T.; Williams, D. H. Synthesis of covalent head-to-tail
dimers of vancomycin. Tetrahedron Lett. 1998, 39, 4917-4920.
(17) Greenwald, R. B.; Gilbert, C. W.; Pendri, A.; Conover, C. D.; Xia,
J .; Martinez, A. Drug delivery systems: Soluble taxol 2′-poly
(ethylene glycol) ester progroup - design and in vitro effective-
ness. J . Med. Chem. 1996, 39, 424-431.
(18) Greenwald, R. B.; Pendri, A.; Conover, C. D.; Zhao, H.; Choe, Y.
H.; Martinez, A.; Shum, K.; Guan, S. Drug delivery systems
employing 1,4-or 1,6-elimination: poly (ethylene glycol) prodrugs
of amine containing compounds. J . Med. Chem. 1999, 42, 3657-
3667.
(19) Greenwald, R. B.; Yang, Y.; Zhao, H.; Conover, C. D.; Lee, S.;
Filpula, D. Controlled release of proteins from their poly-
(ethylene glycol) conjugates: Drug delivery systems employing
1,6-elimination. Bioconjugate Chem. 2003, 14, 395-403.
(20) Lee, S.; Greenwald, R. B.; McGuire, J .; Yang, K.; Shi, C. Drug
delivery systems employing 1,6-elimination: releasable poly-
(ethylene glycol) conjugates of proteins. Bioconjugate Chem.
2001, 12, 163-169.
(21) Choe, Y. H.; Conover, C. D.; Wu, D.; Royzen, M.; Greenwald, R.
B. Anticancer drug delivery systems: N4-acyl poly (ethylene
glycol) prodrugs of ara-C. I. Efficacy in solid tumors. J . Con-
trolled Release 2000, 79, 41-53.
(22) Choe, Y. H.; Conover, C. D.; Wu, D.; Royzen, M.; Gervacio, Y.;
Borowski, V.; Mehlig, M.; Greenwald, R. B. Anticancer drug
delivery systems: Multi-loaded N4-acyl poly (ethylene glycol)
prodrugs of ara-C. II. Efficacy in ascites and solid tumors. J .
Controlled Release 2000, 79, 55-70.
Ack n ow led gm en t. We wish to thank Mr. Clifford
Longley, Dr. Ping Peng, and the Pharmacology Group
of Enzon for their invaluable assistance in preparing
data for this manuscript. Also thanks to Carol Aitken
for her invaluable support.
Refer en ces
(1) Kirby, W. M. M. Vancomycin therapy in severe staphyloccal
infections. Rev. Infect. Dis. 1981, 3, 5236-5239.
(2) Ingerman, M. J .; Santoro, J . Vancomycin a new old agent. Infect.
Dis. Clin. North Am. 1989, 3, 641-651.
(3) Pryka, R. D.; Rodveld, K. A.; Erdman, S. M. An updated
comparison of drug dosing methods. IV. Vancomycin. Clin.
Pharmacokinet. 1991, 20, 463-476.
(4) J ames, J . K.; Palmer, S. M.; Levine, D. P.; Rybac, M. J .
Comparison of conventional dosing versus continuous-infusion
vancomycin therapy for patients with suspected or documented
gram-positive infections. Antimicrob. Agents and Chemother.
1996, 40, 696-700.
(5) Klepser, M. E.; Patel, K. B.; Nicolau, D. P.; Quintiliani, R.;
Nightingale, C. H. Comparison of bactericidal activities of
intermittent and continuous infusion dosing of vancomycin
against methicillin-resistant Staphylococcus aureus and Entero-
coccus faecalis. Pharmacotherapy 1998, 18, 1069-1074.
(6) Wysocki, M. et al. Continuous versus intermittent infusion of
vancomycin in severe Staphyloccal infections: Prospective mul-
ticenter randomized study. Antimicrob. Agents Chemother. 2001,
45, 2460-2467.
(7) Sinkula, S. S.; Yalkowsky, S. H. Rationale for design of biologi-
cally reversible drug derivatives: Prodrugs. J . Pharm. Sci. 1975,
64, 181-210.
(8) Pavlov, A. Y. Preobrazhenskaya. Chemical modification of
glycopeptide antibiotics [VC1]. Russian J . Bioorg. Chem. (Eng.
Version). 1998, 24, 570-587.
(9) Nagarajan, R.; Schabel, A. A.; Occolowitz, J . L.; Counter, F. T.;
Ott, J . L. Synthesis and antibacterial activity of N-acyl vanco-
mycins. J . Antibiot. 1988, XLI, 1430-1438.
(10) Schecter, Y.; Tsubery, H.; Fridkin, M. J Med. Chem. 2002, 45,
4264-4270.
(11) Pang, S. J . Final report on the safety assessment of poly(ethylene
glycol) (PEGs). J . Am. College Toxicol. 1993, 12, 429.
(12) Kannan. R.; Harris, C. M.; Harris, T. M.; Waltho, J . P.; Skelton,
N. J .; Williams, D. H. Function of the amino sugar and
N-terminal amino acid of the antibiotic vancomycin in its
complexation with cell wall peptides. J . Am. Chem. Soc. 1988,
110, 2946-2953.
(23) Yamaoka, T.; Tabata, Y.; Ikada, Y. Distribution and tissue
uptake of poly (ethylene glycol) with different molecular weights
after intravenous administration to mice. J . Pharm. Sci. 1994,
83, 601-606.
J M030202G