Syntheses and studies of multiwarhead siderophore-5-fluorouridine conjugates

Article abstract of DOI:10.1016/S0968-0896(99)00248-5

Siderophores are microbial iron chelating agents that sequester physiologically essential iron for microbes. Conjugation of drugs to siderophores allows use of active iron transport for microbially directed drug delivery. Syntheses and biological studies are described of the first multidrug isocyanurate-based siderophore analogues separately containing one, two, and three 5-fluorouridine (5-FU) derivatives as the drug component. The results indicate that a single siderophore can be used to deliver multiple drugs to target pathogenic microorganisms. Copyright (C) 1999 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(99)00248-5

Bioorganic & Medicinal Chemistry 7 (1999) 3025±3038
Syntheses and Studies of Multiwarhead
Siderophore-5-¯uorouridine Conjugates
Yong Lu and Marvin J. Miller*
Department of Chemistry and Biochemistry, University of Notre Dame, 251 Nieuwland Science Hall, Notre Dame, IN 46556-5670, USA
Received 26 July 1999; accepted 27 August 1999
AbstractÐSiderophores are microbial iron chelating agents that sequester physiologically essential iron for microbes. Conjugation
of drugs to siderophores allows use of active iron transport for microbially directed drug delivery. Syntheses and biological studies
are described of the ®rst multidrug isocyanurate-based siderophore analogues separately containing one, two, and three 5-¯uoro-
uridine (5-FU) derivatives as the drug component. The results indicate that a single siderophore can be used to deliver multiple
drugs to target pathogenic microorganisms. # 1999 Elsevier Science Ltd. All rights reserved.
Introduction
rhodotorolic acid (2), but with the ability to bind ferric
ion stoichiometrically. Previous biological studies indi-
cated that arti®cial siderophore analogue 1 could facil-
itate iron assimilation for a number of microorganisms
including Escherichia coli and Candida albicans. In
addition, isocyanurate analogue 1 was not toxic to
mammals.16,17 Moreover, a series of isocyanurate-based
siderophore conjugates with antimicrobial agents were
synthesized. Preliminary biological studies revealed the
ability of this synthetic siderophore to act as a drug
carrier.17,18 This prompted further investigation into
potential uses of the isocyanurate-based siderophore
analogue (Fig. 2).
Microbial resistance to antibiotics is becoming a major
health care concern especially among people with com-
promised immune systems.1±6 Of several modes of
resistance, the inability of some antibiotics to enter the
cell by the usual passive di€usion process is especially
signi®cant. Thus, the development of methods to facil-
itate active transport of antibiotics into microbial cells is
an important therapeutic goal. One attractive approach
to facilitate active transport of drugs into microbes is
through conjugation to known molecules which the cell
must assimilate for survival.7 Siderophores are natural
iron chelators utilized by many microbes for sequester-
ing and cellular transport of physiologically essential
iron. In principle, siderophore-mediated active trans-
port may overcome the permeability problem by smug-
gling the drugs into microbial cells via microbial iron
transport systems. Selective recognition and delivery of
the drugs to target microbes may serve to improve the
ecacy of the antimicrobial agents. The therapeutic
potential of siderophore±drug conjugates (Fig. 1)8±10
has been reviewed before.11±15
Results and Discussion
Design of conjugates
The synthetic siderophore 1 has an isocyanurate plat-
form with three attached iron-chelating hydroxamates.
Replacement of the acetyl groups on the three hydro-
xamates with succinoyl groups should provide com-
pound 3, which has the potential for attachment of
three drugs to the same siderophore component. The
three drugs can be the same or even di€erent creating a
We have previously reported the syntheses and studies
of isocyanurate-based siderophores.16,17 Trihydrox-
amate derivative 1 was initially designed in our labora-
tory to mimic the natural dihyroxamate siderophore
multiwarhead siderophore±drug conjugate. For
a
drug delivery system, multiple drug conjugates might
increase the delivery eciency since every single carrier
is envisioned to carry more than one drug. On the other
hand, the multiwarhead conjugates are relatively large
compared to the parent siderophore analogue 1. Thus,
a potential concern was whether the multiwarhead
* Corresponding author. Tel.: +1-219-631-7058; fax: +1-219-631-
6652; e-mail: marvin.j.miller.2@nd.edu
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00248-5

3026
Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
siderophore±5-FU conjugates, use of di€erent amino
acid spacers was studied and valine-containing con-
jugates were found to be less prone to ester hydrolysis
(half-life ꢀ241 h) than other amino acid conjugates
under assay conditions18 (Fig. 4).
Figure 1. General siderophore±drug conjugate.
Synthesis of conjugates
The synthesis of the isocyanurate-based siderophore
analogues was initiated by the preparation of a pro-
tected hydroxamate side chain precursor as outlined
in Scheme 1. After treatment of N-trichloroethoxy-
carbonyl (troc) O-benzyl hydroxylamine (5) with NaH,
the resulting sodium salt was added dropwise to an
excess of 1,4-dibromobutane (6). This method provided
the desired product 7 and bisalkylated product 8 in a
ratio of 11:1. The following steps in the sequence paral-
leled our earlier synthesis of 10.17 Construction of the
core structure of the isocyanurate-based siderophore
analogue 10 was completed by treatment of cyanuric
acid (9) with NaH followed by addition of bromide 7
(Scheme 1).
Figure 2.
With tris-troc-protected analogue 10 in hand, sequential
removal of troc protecting groups was explored in the
hopes of obtaining the isocyanurate core with di€eren-
tially functionalized handles. Complete acylation of 10
in the presence of an excess of freshly activated Zn dust
and succinic anhydride provided tris-succinate 11 in
95% yield (Scheme 2). In order to sequentially remove
the troc protecting groups, the reaction rate needed to
be carefully controlled. Several di€erent reductive
methods for removal of the troc protecting group are
reported in the literature.21,22 With Zn/AcOH, the
reaction rate is highly dependent on the acidity of the
reaction medium.22 A method involving fast initiation
of the reaction followed by an immediate quench was
investigated previously in our laboratory.17 It was
found that the concentration of Zn dust in the solution
greatly in¯uenced the reaction rate. When a large excess
of Zn was used in the reaction, all three troc groups
were removed quickly and almost simultaneously.
When Zn was quite dilute in the reaction mixture, the
troc groups were removed sequentially. Under carefully
optimized conditions, the mono-, bis- and tris-succinate
(11, 12 and 13) were obtained as major products,
respectively. Careful control of Zn concentration,
reaction temperature, reaction time and the acidity of
reaction media were all important factors. As shown in
Scheme 2, treatment of 10 with three equivalents of Zn
and succinic anhydride in THF containing 10% AcOH
at 55±60ꢁC followed by quenching of the reaction
within 1 h provided optimal distribution of mono- and
bis-succinate siderophore analogues. The unreacted
conjugates could still be recognized and actively trans-
ported by microbes. However, previous studies of side-
rophore±drug conjugates showed that the siderophore
component appears to still be recognizable and utilized
despite the fact that, in some cases, the drug component
of the conjugate is much larger than the siderophore
component19 (Fig. 3).
Our ®rst choice for drug attachment was the anti-
microbial agent 5-¯uorouridine (5-FU), an active meta-
bolite of the commonly used active antifungal agent
5-¯uorocytosine (5-FC).20 Valine was chosen as a spacer
between the isocyanurate core and 5-FU for two rea-
sons: (1) Val-5-FU is not only more active against fungi
than 5-FU itself, but also shows strong antibacterial
activity. (2) In earlier studies with ornithine-based
Figure 3.
Figure 4. Valine-5-¯urouridine (val-5-FU).

Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
3027
Scheme 1.
Scheme 2.
starting material could then be recycled, adding to the
utility of this modi®ed procedure. By increasing the
amount of Zn to six equivalents and stirring the reac-
tion mixture at 35±40ꢁC for 7±8 h, bis-succinate 12 was
obtained in a 71% yield. The three di€erent products
and the two starting materials were easily separated by
radial chromatography. The ability to sequentially
remove the troc protecting groups allowed for the
potential introduction of three di€erent drugs to a single
iron complex. With these siderophore analogues in
hand, the preparation of isocyanurate-based side-
rophore±multidrug conjugates was initiated.
obtained as the major product. According to the litera-
ture precedent,23 HOAt esters are more reactive than
the esters of commonly used HOBt (1-hydroxy-7-ben-
zotriazole). Although precise mechanistic details have
not been established, the reactivity increase is pre-
sumably due to assistance from internal hydrogen
bonding of the incoming nucleophiles by the nitrogen
atom of pyridine residue in HOAt. The optimal cou-
pling eciency was found to be use of EDC alone.
The synthesis of a tris-5-FU conjugate 19 was explored
next (Scheme 4). Deprotection of the Cbz group of 14 in
the presence of one equivalent of p-TsOH provided the
corresponding amine salt, which was immediately acy-
lated with tris-succinate siderophore 13 using HOAt
and EDC. The reaction was sluggish and problematic
without the use of p-TsOH in the hydrogenation step,
therefore one equivalent of p-TsOH was required for
ecient deprotection. The coupling reaction was found
To synthesize the val-5FU component 14, EDC-medi-
ated ester®cation of 5-FU with Cbz-l-valine with dif-
ferent coupling additives was explored (Scheme 3).
HOAt (1-hydroxy-7-azabenzotriazole) was found to be
overly e€ective as a coupling additive. When HOAt
was used, the interesting bis-val-5-FU adduct 15 was

3028
Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
Scheme 3.
Scheme 4.
to be dependent on the acidity of the reaction media.
When a catalytic amount of DMAP was used, the major
product was bis-conjugate 18, however, when a full
equivalent of DMAP was used, tris-conjugate 19 was
obtained as the major product.
bis-troc analogue 20. It was then discovered that the Zn/
AcOH conditions employed previously for troc protect-
ing group removal caused decomposition of the
conjugate. To overcome this problem, alternative meth-
odology was utilized for troc removal.24,25 Fortunately,
10% Cd±Pb couple was found to be an ecient redu-
cing reagent in a pH 7, 1 M aq NH4OAc bu€er solution
providing 17 in 85% yield after acylation with succinic
anhydride (Scheme 5).
Synthesis of mono-val-5FU conjugate (17) was attemp-
ted by an alternative route (Scheme 5). The ®rst target
was the mono-val-5FU conjugate as the protected

Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
3029
Scheme 5.
The benzyl protecting groups of all of the coupled pro-
ducts were removed by hydrogenolysis to obtain the
desired conjugates in their free hydroxamate forms
(Scheme 6).
utilize siderophores under these conditions, as expected.
Thus, the bacteria recognized each conjugate as a source
of nutritional iron, but in the process of actively assim-
ilating it, the bacteria also assimilated the antibiotic and
in e€ect committed suicide. The multiwarhead con-
jugates were not only active against the bacteria, but
also more potent than mono conjugate 21 (Fig. 8). In
fact, the tris-val-5-FU conjugate 23 was actually more
potent than val-5-FU itself (Fig. 8). This may indicate
that E. coli X580 still can recognize and actively trans-
port these multiwarhead conjugates as planned.
Biological studies
To determine the antimicrobial activities of side-
rophore±drug conjugates, di€erent assay methods have
been employed. Previous studies using only the standard
broth or agar dilution assays to determine the minimum
inhibitory concentration (MIC) were found to be mis-
leading.17 Thus, kinetic growth assays were performed
on compounds 21, 22 and 23. Both ferri (with stoichio-
metric Fe3+ added) and deferri forms of each of the
conjugates were used for this study. All of the con-
jugates were tested in regular Luria broth and also in
an iron-de®cient medium. If the regular Luria broth
was pretreated with ethylenediamine bis(O-hydroxy-
phenylacetic acid) (EDDA) before the test, it was con-
sidered to be an iron-de®cient environment for the
microbes.17
To compare the delivery eciency of multiwarhead
conjugates 22 and 23 with mono conjugate 21, further
biological assays were performed (Figs 9 and 10). Tris-
conjugate 23, bis-conjugate 22, and mono-conjugate 21
were tested at 1.25, 1.88, and 3.75 mM, respectively, in
the high concentration assay. However, the total con-
centration of val-5-FU from all conjugates was equal to
3.75 mM since one molecule of 26 possesses three mole-
cules of val-5-FU. As shown in Figure 9, the anti-
bacterial activity of all three conjugates are very close
to each other under iron sucient conditions with
tris-conjugate 23 being only slightly more potent than
other conjugates. However, tris-conjugate 23 was more
active than the mono- (21) and bis-conjugates (22)
under iron-de®cient conditions (Fig. 10). The con-
centration of mono-conjugate 21 (3.75 mM) was as much
as three times that of tris-conjugates (1.25 mM).
Although the concentration of val-5FU in each case
Synthetic conjugates 21, 22 and 23 were subjected to
studies of antibacterial activity against E. coli X580 at
10, 5 and 2.5 mM. At these concentrations, all of the
val-5-FU conjugates displayed certain activity against
E. coli X580 (Figs 5±7). All the tested conjugates
showed much stronger activity under iron-de®cient
conditions indicating the bacteria will more actively
Scheme 6.

3030
Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
Figure 5. Growth curves of E. coli X580 in Luria broth in the presence of tris-val-5FU conjugate 23.
Figure 6. Growth curves of E. coli X580 in Luria broth in the presence of bis-val-5FU conjugate 22.
could be considered the same (3.75 mM), higher activity
of the tris-conjugate suggests that multiwarhead side-
rophore±drug conjugate 23 is a more ecient drug
delivery system than the mono siderophore±drug con-
jugate 21, or intracellular release is more e€ective.
to have equal e€ective concentrations of val-5-FU. The
results are summarized in Figures 15 and 16. The mul-
tiwarhead conjugates (22 and 23) showed stronger
activity under all conditions. Under iron-de®cient con-
ditions, the multiwarhead conjugates were much more
potent than mono-conjugate (21), which was consistent
with previous results.
The antifungal activity of conjugates 21, 22 and 23
against C. albicans was also determined by kinetic
growth assays in Luria broth cultures. Results similar to
the antibacterial test were obtained and are summarized
in Figures 11±14. All of the conjugates showed some
delay of growth of fungi and multiwarhead side-
rophore±val-5FU conjugates were more potent than the
mono siderophore±val-5-FU conjugate 21.
The stability of siderophore-amino acid-5-FU con-
jugates under the assay conditions was studied pre-
viously in our laboratory.18 As in related studies with
isocyanurate-based siderophore±val-5-FU conjugate 24
(Fig. 17), a hydrolysis study employing 19F NMR spec-
troscopy was used to determine the relative stability of
conjugate 24 under bioassay conditions. Conjugate 24
showed less than 5% hydrolysis even after 19 h, corre-
sponding to a half-life on the order of 296 h. Appar-
ently, this result reveals that the test compounds are
The antifungal activity of the conjugates was also further
investigated. The conjugates were tested at three con-
centrations respectively, each of which was considered

Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
3031
Figure 7. Growth curves of E. coli X580 in Luria broth in the presence of mono-val-5FU conjugate 21.
Figure 8. Growth curves of E. coli X580 in the presence of 23, 22 and 21 and val-5FU at 10 mM under regular conditions.
stable under the assay conditions, indicating that the
observed results are not due to spontaneous decom-
position of the conjugates in the culture broth.
conjugates are being carried out in our laboratory and
will be reported in due course.
In conclusion, the isocyanurate-based siderophore±
val-5-FU conjugates again displayed promising anti-
microbial activities demonstrating the therapeutic
potential of siderophore±drug conjugates in combating
pathogenic infections. The multiwarhead conjugates
were more potent against microbes than the regular
mono conjugate supporting the assumption that the
multiwarhead siderophore±drug conjugate is a more
ecient drug delivery system, at least in the preliminary
studies completed. The ability of sequential introduction
of the linkage functionality may lead to the synthesis of
multiwarhead conjugates with three di€erent drugs. The
biological behavior of such conjugates should be very
interesting. Further synthetic and biological studies of
the isocyanurate-based multiwarhead siderophore±drug
Experimental
General methods
General methods and instruments used have been
described previously.17 The term `dried' refers to the
drying of an organic layer over anhydrous sodium sul-
fate. All reactions were performed under nitrogen
atmosphere. Compounds 7, 10 and 13 have been pre-
viously described.17
[Mono[N-succinoyl]bis[N-(trichloroethoxy)] (benzyloxy)-
amino]butyl] isocyanurate (11). Compound 1017 (0.5 g,
0.421 mmol) was dissolved in 11 mL of a 1:10 mixture of
AcOH and THF. Freshly activated zinc dust (82.1 mg,

3032
Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
1.263 mmol) and succinic anhydride (126.3 mg,
1.263 mmol) were added to the resulting solution. The
reaction mixture was stirred 40 min±1 h (monitored by
TLC) at 45±50ꢁC (water bath). Then it was diluted with
cold EtOAc and residual Zn dust was removed by ®l-
tration. The solvent was evaporated and the residue was
taken up into EtOAc. The EtOAc solution was washed
with water several times and dried. After ®ltration and
evaporation of the solvent, the residue was chromato-
graphed over a Chromatotron1 silica gel plate. Elution
with a mixture of CH2Cl2, i-PrOH, and AcOH (100:2:1,
100:5:1, 100:10:1 gradient) provided 11 (70.2 mg, 15%),
148.80, 173.12, 177.15; IR (neat) 1460, 1690, 1700, 1720
1
cm
;
MS (FAB) m/e 1112.67 (MH+), 1134.54
(MNa+).
[[Bis[N-succinoyl]mono[N-(trichloroethoxy)] N-(benzyloxy)-
amino]butyl] isocyanurate (12). Compound 10 (0.5 g,
0.421 mmol) was dissolved in 10 mL of a 1:1 mixture of
AcOH and THF. Freshly activated zinc dust (164.2 mg,
2.526 mmol) and succinic anhydride (252.6 mg,
2.526 mmol) were added to the resulting solution. The
reaction mixture was stirred 7±10 h (monitored by TLC)
at 35±40ꢁC. Then it was diluted with cold EtOAc and
residual Zn dust was removed by ®ltration. The solvent
was evaporated and the residue was taken up into
EtOAc. The EtOAc solution was washed with water
several times and dried. After ®ltration and evaporation
of the solvent, the residue was chromatographed over a
Chromatotron1 silica gel plate. Elution with a mixture
1
12 (17.4 mg, 4%) and 10 (410 mg, 82%). 11: H NMR
(CDCl3) d 1.45±1.75 (m, 12H), 2.55±2.75 (m, 4H), 3.62±
3.72 (m, 6H), 3.75±3.87 (m, 6H), 4.82 (s, 6H), 4.92 (s,
4H), 7.36 (m, 15H), 9.6 (br s, 3H); 13C NMR (CDCl3) d
24.14, 24.87, 27.12, 28.62, 42.46, 49.11, 75.08, 77.20,
128.49, 128.71, 128.79, 129.00, 129.17, 129.47, 134.73,

Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
3033
Figure 11. Growth curves of C. albicans in the presence of tris-val-5FU conjugate 23.
Figure 12. Growth curves of C. albicans in the presence of bis-val-5FU conjugate 22.
of CH2Cl2, i-PrOH, and AcOH (100:2:1, 100:5:1,
100:10:1 gradient) provided 11 (70.2 mg, 15%), 12
(309.1 mg, 71%), 13 (28.32 mg, 7%) as colorless clear
solution was washed with 0.5 M HCl, 0.3 M NaHCO3,
water and brine. The organic layer was then dried, ®l-
tered and evaporated. The residue was chromato-
graphed over silica gel using a Chromatotron1 with
CH2Cl2: i-PrOH (95:5). This a€orded 14 (59.51 mg,
1
oils. 12: H NMR (CDCl3) d 1.45±1.75 (m, 12H), 2.55±
2.75 (m, 8H), 3.48±3.72 (m, 6H), 3.75±3.87 (m, 6H), 4.83
(s, 6H), 4.92 (s, 3H), 7.34±7.39 (m, 15H); 13C NMR
(CDCl3) d 23.63, 24.51, 26.92, 28.31, 42.19, 44.58, 75.98,
77.19, 128.44, 128.71, 128.95, 133.98, 148.64, 173.12,
177.15; IR (neat) 1460, 1690, 1700, 1720 cm 1; MS
(FAB) m/e 1035.30 (M+), 1057.37 (MNa+).
62%) as a white solid. Mp: 64±67ꢁC H NMR (CDCl3)
1
d 0.91 (d, 3H), 0.98 (d, 3H), 2.03±2.23 (m, 1H), 3.85±
4.55 (m, 7H), 5.01±5.12 (m, 2H), 5.46 (d, 1H), 5.98 (d,
1H), 7.26±7.38 (m, 5H), 8.02 (d, 1H); 13C NMR
(CDCl3): d 17.33, 17.54, 19.01, 30.95, 59.48, 63.94,
67.22, 69.88, 74.39, 77.22, 81.74, 89.87, 128.05, 128.19,
128.25, 128.51, 135.91, 142.51, 149.71, 156.55, 157.53,
171.88, 176.18; IR 1580, 1700 (br), 3450 (br) cm 1; MS
(FAB) m/e calcd 496.1712 (MH+), found 496.1738
(MH+).
50 -O-(N-(Benxyloxycarbonyl)-L-valinyl)-5-¯uorouridine
(14). Cbz-valine (50 mg, 0.2 mmol) was dissolved in
4 mL of THF, then 5-FU (52.1 mg, 0.2 mmol) was
added to the ¯ask, followed by 4 mL of DMF to dis-
solve the 5-FU. Then EDC (38.1 mg, 0.2 mmol) and
DMAP (24.2 mg, 0.2 mmol) were added. The reaction
mixture was stirred at rt overnight. The reaction
mixture was diluted with EtOAc (75 mL). The EtOAc
[Bis[N-(trichloroethoxy)carbonyl]mono[N-succinoyl] (benzyl-
oxy)amino butyl] isocyanurate (20). To a solution of
Cbz-valine-5-FU (31.2 mg, 0.063 mmol) in 5 mL of

3034
Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
Figure 13. Growth curves of C. albicans in the presence of mono-val-5FU conjugate 21.
Figure 14. Growth curves of C. albicans in the presence of 21, 22 and 23 and val-5FU.
deoxygenated MeOH was added p-TsOH (12 mg,
0.063 mmol) and 10% Pd±C (6 mg, 20% w/w). The sus-
pension was stirred under 1 atm of H2 for 50 min after
which TLC analysis indicated that the reaction was
complete. The catalyst was removed by ®ltration and
the solvent was evaporated to a€ord the p-TsOH salt of
the amine. The crude salt was dissolved in 5 mL of
anhydrous DMF. A solution of 11 (70 mg, 0.063 mmol)
in 5 mL of DMF was added, quickly followed by HOAt
(8.6 mg, 0.063 mmol), EDC (12 mg, 0.063 mmol), and
DMAP (7.6 mg, 0.063 mmol). The reaction mixture was
stirred at rt overnight, then diluted with EtOAc (75 mL).
The EtOAc solution was washed with 0.5 M HCl, 0.3 M
NaHCO3 water and brine. Then the solvent was dried,
®ltered and evaporated. The residue was chromato-
graphed over a Chromatotron1 silica gel plate eluting
with CH2Cl2: i-PrOH (95:5). This a€orded compound
0.92±0.99 (d, 6H), 1.45±1.75 (m, 12H), 2.01±2.23 (m,
1H), 2.55±2.75 (m, 4H), 3.62±3.86 (m, 12H), 4.15±4.48
(m, 7H), 4.81 (s, 6H), 4.89 (s, 4H), 5.18 (d, 1H), 5.91 (d,
1H), 7.32±7.42 (m, 15H), 8.08 (d, 1H); 13C NMR
(CDCl3) d 17.98, 18.97, 20.38, 23.88, 24.68, 24.87, 27.64,
29.92, 30.69, 42.46, 44.73, 58.07, 61.78, 69.42, 74.15,
76.19, 76.25, 77.20, 81.63, 91.05, 125.3, 128.64, 128.92,
129.13, 134.09, 134.21, 139.12, 142.08, 148.90, 149.13,
157.05, 171.20, 172.43, 172.83, 176.82; IR 1460, 1580,
1720 (br) cm 1; MS (ES) m/e 1478.01 (MNa+).
10% Cadmium-lead couple. Lead oxide (PbO, 108 mg,
0.49 mmol) was dissolved in 5 mL of 50% aq AcOH,
and the solution was slowly added to a vigorously stir-
red suspension of Cd dust (100 mesh, 546 mg, 4.9 mmol)
in deionized water (10 mL). The Cd darkened as Pb
deposited on its surface, and formed clumps that were
broken up with a glass rod. The Cd±Pb couple was
1
20 (64.2 mg, 70%) as a clear oil. H NMR (CDCl3) d

Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
3035
Figure 15. Growth curves of C. albicans in the presence of 21, 22 and 23 at regular condition.
Figure 16. Growth curves of C. albicans in the presence of 21, 22 and 23 at iron de®cient conditions.
®ltered, washed with water, then acetone, vacuum dried,
ground, and stored in a closed vessel.
water, brine, then dried, ®ltered and concentrated. The
crude oil was puri®ed using a Chromatotron1 silica gel
plate eluting with CH2Cl2: i-PrOH:MeOH:AcOH
(90:5:5:1). This gave compound 17 as a clear oil
[Mono[N-succinoyl [(valinyl)-5-¯uorouridine]-bis[N-succi-
noyl] (benzyloxy)amino butyl] isocyanurate (17). Cd±Pb
couple (24.8 mg, 0.206 mmol) was added to a vigorously
stirring mixture of compound 20 (20 mg, 0.0137 mmol),
THF (4 mL) and 1 M NH4OAc (1 mL). Cloudiness
slowly developed. When the reaction was complete
(TLC monitored), the solids were ®ltered and washed
with water and THF. The ®ltrate was diluted with
EtOAc. Then the EtOAc layer was washed with brine,
dried, ®ltered and evaporated to a€ord a slight-yellow
oil. The oil was dissolved in THF (4 mL) and DMF
(4 mL). Then succinic anhydride (4.11 mg, 0.0411 mmol)
and a catalytic amount of DMAP were added. The
reaction mixture was stirred at rt overnight, then diluted
with EtOAc. The EtOAc solution was washed with
1
(15.2 mg, 85%). H NMR (CDCl3) d 0.92±0.99 (d, 6H),
1.45±1.75 (m, 6H), 2.01±2.23 (m, 1H), 2.55±2.75 (m,
12H), 3.62±3.86 (m, 12H), 4.15±4.48 (m, 7H), 4.81 (s,
6H), 5.18 (d, 1H), 5.91 (d, 1H), 7.32±7.42 (m, 15H), 8.08
(d, 1H); 13C NMR (CDCl3) d 17.98, 18.97, 20.38, 23.88,
24.68, 24.87, 27.64, 29.92, 30.69, 42.46, 44.73, 58.07,
61.78, 69.42, 74.15, 76.19, 76.25, 77.20, 81.63, 91.05,
125.3, 128.64, 128.92, 129.13, 134.09, 134.21, 139.12,
142.08, 148.90, 149.13, 157.05, 171.20, 172.43, 172.83,
176.82; IR 1580, 1700 (br), 3450 (br) cm 1; MS (ES) m/e
1327.34 (MNa+).
[Bis[N-succinoyl [(valinyl)-5-¯uorouridine]mono[N-succi-
noyl] (benzyloxy)amino butyl] isocyanurate (18). To a

3036
Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
Figure 17.
solution of Cbz-valine-5-FU (31.2 mg, 0.063 mmol) in
5 mL of deoxygenated MeOH was added p-TsOH
(12 mg, 0.063 mmol) and 10% Pd±C (6 mg, 20% w/w).
The suspension was stirred under 1 atm of H2 for 40 min
after which time TLC analysis indicated that the
reaction was complete. The catalyst was removed by
®ltration and the solvent was evaporated to a€ord the
p-TsOH salt of the amine. The crude salt was dissolved
in 5 mL of anhydrous DMF. A solution of 13 (20.2 mg,
0.021 mmol) in 5 mL of DMF was added, quickly fol-
lowed by HOAt (8.6 mg, 0.063 mmol), EDC (12 mg,
a€ord compound 19 (18.8 mg, 45%) and compound 18
1
(12.11 mg, 35%) as clear oils. 19: H NMR (CDCl3) d
0.92±0.99 (d, 18H), 1.45±1.75 (m, 12H), 2.01±2.23 (m,
3H), 2.55±2.75 (m, 12H), 3.62±3.86 (m, 12H), 4.15±4.48
(m, 21H), 4.81 (s, 6H), 5.18 (d, 3H), 5.91 (d, 3H), 7.32±
7.42 (m, 15H), 8.08 (d, 3H); 13C NMR (CDCl3) d 17.98,
18.97, 20.38, 23.88, 24.68, 24.87, 27.64, 29.92, 30.69,
42.46, 44.73, 58.07, 61.78, 69.42, 74.15, 76.19, 76.25,
77.20, 81.63, 91.05, 125.3, 128.64, 128.92, 129.13,
134.09, 134.21, 139.12, 142.08, 148.90, 149.13, 157.05,
171.20, 172.43, 172.83, 176.82; IR 1460, 1580, 1700 (br),
3450 (br) cm 1, MS (ES) m/e 2014.88 (MNa+).
0.063 mmol), and
a catalytic amount of DMAP
(<1 mg). The reaction mixture was stirred at rt over-
night, then diluted with EtOAc (75 mL). The EtOAc
solution was washed with 0.5 M HCl, 0.3 M NaHCO3,
water and brine. Then the solvent was dried, ®ltered,
and evaporated. The residue was chromatographed
though a Biotage Flash 40 silica gel cartridge eluting
with CH2Cl2: i-PrOH:AcOH (95:5:1 to 90:10:1 gradient)
to a€ord compound 17 (4.1 mg, 15%), 18 (20 mg, 58%)
and 19 (4.2 mg, 10%) as clear oils. 18: 1H NMR
(CDCl3) d 0.92±0.99 (d, 12H), 1.43±1.72 (m, 12H), 2.01±
2.23 (m, 2H), 2.55±2.75 (m, 12H), 3.62±3.86 (m, 12H),
4.15±4.48 (m, 14H), 4.82 (s, 6H), 5.18 (d, 2H), 5.91 (d,
2H), 7.32±7.42 (m, 15H), 8.08 (d, 2H); 13C NMR
(CDCl3) d 17.98, 18.97, 20.38, 23.88, 24.68, 24.87, 27.64,
29.92, 30.69, 42.46, 44.73, 58.07, 61.78, 69.42, 74.15,
76.19, 76.25, 77.20, 81.63, 91.05, 125.3, 128.64, 128.92,
129.13, 134.09, 134.21, 139.12, 142.08, 148.90, 149.13,
157.05, 171.20, 172.43, 172.83, 176.82; IR 1460, 1580,
1700 (br), 3450 (br) cm 1, MS (ES) m/e 1670.66 (MNa+).
[Mono[N-succinoyl [(valinyl)-5-¯uorouridine]bis[N-succi-
noyl] hydroxyamino] butyl] isocyanurate (21). To a solu-
tion of 17 (35 mg, 0.027 mmol) in 6 mL of MeOH was
added 7 mg of 10% Pd±C. This mixture was exposed to
H2 at atmospheric pressure for 4 h (monitored by TLC).
The catalyst was removed by ®ltration and the solvent
was removed by evaporation under vacuum. This a€or-
1
ded pure 21 (27.9 mg, 99%) as a foamy solid. H NMR
(CDCl3) d 0.92±0.99 (d, 6H), 1.45±1.75 (m, 6H), 2.01±
2.23 (m, 1H), 2.55±2.75 (m, 12H), 3.62±3.86 (m, 12H),
4.15±4.48 (m, 7H), 5.18 (d, 1H), 5.91 (d, 1H), 8.08 (d,
1H); 1C NMR (CDCl3) d 17.98, 18.97, 20.38, 23.88,
24.68, 24.87, 27.64, 29.92, 30.69, 42.46, 44.73, 58.07,
61.78, 69.42, 74.15, 76.19, 76.25, 77.20, 81.63, 91.05,
125.3, 134.21, 139.12, 142.08, 148.90, 149.13, 157.05,
171.20, 172.43, 172.83, 176.82; IR (neat) 1460, 1700
(br), 3450 (br) cm 1; MS (ES) m/e 1057.21 (MNa+).
[Bis[N-succinoyl [(valinyl)-5-¯uorouridine]mono[N-succi-
noyl] hydroxyamino] butyl] isocyanurate (22). To a solu-
tion of 18 (40 mg, 0.024 mmol) in 6 mL of MeOH was
added 8 mg of 10% Pd±C. This mixture was exposed to
H2 at atmospheric pressure for 4 h (monitored by TLC).
The catalyst was removed by ®ltration and the solvent
was removed by evaporation under vacuum. This a€or-
Tris[N-succinoyl [(valinyl)-5-¯uorouridine] (benxyloxy)-
amino]butyl] isocyanurate (19). To a solution of Cbz-
valine-5FU (31.2 mg, 0.063 mmol) in 5 mL of deoxy-
genated MeOH was added p-TsOH (12 mg, 0.063 mmol)
and 10% Pd±C (6 mg, 20% w/w). The suspension was
stirred under 1 atm of H2 for 40 min after which TLC
analysis indicated that the reaction was complete. The
catalyst was removed by ®ltration and the solvent was
evaporated to a€ord the p-TsOH salt of the amine. The
crude salt was dissolved in 5 mL of anhydrous DMF. A
solution of 13 (20 mg, 0.021 mmol) in 5 mL DMF was
added, quickly followed by HOAt (8.6 mg, 0.063 mmol),
EDC (12 mg, 0.063 mmol) and DMAP (15.3 mg,
0.126 mmol). The reaction mixture was stirred at rt
overnight, then diluted with EtOAc (75 mL). The
EtOAc was washed with 0.5 M HCl, 0.3 M NaHCO3,
water and brine. Then the solvent was dried, ®ltered and
evaporated. The residue was chromatographed through
a Biotage Flash 40 silica gel cartridge eluting with
CH2Cl2:i-PrOH:AcOH (95:5:1 to 90:10:1 gradient) to
1
ded pure 22 (23.4 mg, 99%) as a foamy solid. H NMR
(CDCl3) d 0.92±0.99 (d, 12H), 1.45±1.75 (m, 12H), 2.01±
2.23 (m, 2H), 2.55±2.75 (m, 12H), 3.62±3.86 (m, 12H),
4.15±4.48 (m, 14H), 5.18 (d, 2H), 5.91 (d, 2H), 8.08 (d,
2H); 13C NMR (CDCl3) d 17.98, 18.97, 20.38, 23.88,
24.68, 24.87, 27.64, 29.92, 30.69, 42.46, 44.73, 58.07,
61.78, 69.42, 74.15, 76.19, 76.25, 77.20, 81.63, 91.05,
125.3, 134.21, 139.12, 142.08, 148.90, 149.13, 157.05,
171.20, 172.43, 172.83, 176.82; IR 1460, 1700 (br), 3450
(br) cm 1; MS (ES) m/e 1378.37 (MNa+), 1400.27
(MNa+).
[Tris[N-succinoyl [(valinyl)-5-¯uorouridine] hydroxyamino
butyl] isocyanurate (23). To a solution of 19 (40 mg,

Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
3037
0.02 mmol) in 6 mL of MeOH was added 8 mg of 10%
Pd±C. This mixture was exposed to H2 at atmospheric
pressure for 4 h (monitored by TLC). The catalyst was
removed by ®ltration and the solvent was removed by
evaporation under vacuum. This a€orded pure 23
(34.4 mg, 99%) as a clear oil. 1H NMR (CDCl3) d 0.92±
0.99 (d, 18H), 1.45±1.75 (m, 18H), 2.01±2.23 (m, 3H),
2.55±2.75 (m, 12H), 3.62±3.86 (m, 12H), 4.15±4.48 (m,
21H), 5.18 (d, 3H), 5.91 (d, 3H), 8.08 (d, 3H); 13C NMR
(CDCl3) d 17.98, 18.97, 20.38, 23.88, 24.68, 24.87, 27.64,
29.92, 30.69, 42.46, 44.73, 58.07, 61.78, 69.42, 74.15,
76.19, 76.25, 77.20, 81.63, 91.05, 125.3, 134.21, 139.12,
142.08, 148.90, 149.13, 157.05, 171.20, 172.43, 172.83,
176.82; IR 1460, 1700 (br), 3450 (br) cm 1; MS (ES) m/e
1721.41 (MH+), 1743.58 (MNa+).
The 96-well plate containing the cultures was then
inoculated at 37ꢁC for 24 h by the Molecular Devices
Thermomax1 microplate reader. The culture turbity
was read at 650 nm automatically by the reader every
30 min.
Acknowledgements
Support of this research by the National Institutes of
Health (GM25845 and AI30988) is gratefully acknowl-
edged. We thank Dr. William Bogess and Ms. Nonka
Sevova for their assistance in the MS study, Mr. Donald
Schi€erl and Dr. Jaroslav Zajicek for their assistance in
the NMR study.
Biological procedures for antibacterial and antifungal
testing
References
The assay used to determine the antifungal and anti-
bacterial activity of conjugates 21, 22 and 23 was per-
formed in Luria broth. The yeast strain C. albicans
(ATCC 48130) was obtained from the American Type
Culture Collection, Rockville, MD, and the parent
strain E. coli X580 was the generous gift of Eli Lilly &
Co., Indianapolis, IN. Both strains were stored at
78ꢁC in culture tubes in a 1:1 mixture of Luria broth
culture and sterile d-lactose-glycerol solution.
1. Travis, J. Science 1994, 264, 360.
2. Davies, J. Science 1994, 264, 375.
3. Nikaido, H. Science 1994, 264, 382.
4. Spratt, B. G. Science 1994, 264, 388.
5. Balter, M. Science 1998, 282, 1623.
6. Huang, H. F.; Chopra, R.; Verdine, G. L. Science 1998,
282, 1669.
7. Miller, M. J.; McKee, J. A.; Minnick, A. A.; Dolence, E. K.
Biol. Met. 1991, 4, 62.
8. Neilands, J. B.; Konopka, K.; Schwyn, B.; Coy, M.; Fran-
cis, R. T.; Paw, B. H.; Bagg, A. Iron Transport in Microbes,
Plants and Animals. VCH: Weinheim, 1987.
9. Neilands, J. B. J. Biol. Chem. 1995, 270, 26723.
10. Rogers, H. J. In Iron Transport in Microbes; Winkelmann,
G.; van der Helm, D.; Neilands, J. B., Eds.; VCH: Weinheim,
1987; p. 223.
11. Miller, M. J. Chem. Rev. 1989, 89, 1563.
12. Miller, M. J.; Malouin, F.; Dolence, E. K.; Gasparski, C.
M.; Ghosh, M.; Guzzo, P. R.; Lotz, B. T.; McKee, J. A.;
Minnick, A. A.; Teng, M. In Recent Advances in the Chemistry
of Anti-infective Agents; Bentley, P. H.; Ponsford, R., Eds.;
The Royal Society of Chemistry: Cambridge, 1993; Vol. 119,
p. 135.
13. Miller, M. J.; Malouin, F. Acc. Chem. Res. 1993, 26, 241.
14. Miller, M. J.; Malouin, F. In The Development of Iron
Chelators for Clinical Use; Bergeron, R. J., Brittenham, G. M.,
Eds.; CRC: Ann Arbor, 1994; p. 275.
15. Miller, M. J.; Darwish, I.; Ghosh, A.; Ghosh, M.; Hansel,
J.-G.; Hu, J.; Niu, C.; Ritter, A.; Scheidt, K.; Suling, C.; Sun,
S.; Zhang, D.; Budde, A.; De Clercq, E.; Leong, S.; Malouin,
F.; Moellmann, U. In Anti-infectives: Recent Advances in
Chemistry and Structure±Activity Relationships; Bentley, P. H.;
O'Hanlon, P. J., Eds.; The Royal Society of Chemistry:
Cambridge, 1997; p. 116.
All inorganic compounds were obtained from Fisher.
EDDA, the ferric iron chelator used to deferrate the
culture medium, was obtained from Sigma and was
deferrated using Rogers' procedure,26 and a stock solu-
tion27 (50 mg/mL) was prepared immediately prior to
use. Dehydrated tryptone and yeast extract were all
purchased from Fisher. Luria broth was prepared by the
procedure of Miller.28 The ferric iron chelator EDDA
was added by sterile ®ltration through an Acrodisc 4192
sterile assembly (pore size, 0.2 mm; Fisher) to a slightly
concentrated culture broth. The ®nal culture volume
was typically 100 mL.
Seed cultures were prepared by adding 0.20 mL aliquot
of thawed bacteria or fungi in a culture tube containing
sterile lactose-glycerol to 100 mL of sterile Luria broth.
After overnight incubation at 37ꢁC and 300 rpm, two
10 mL aliquots of the turbid seed culture were trans-
ferred to corresponding control and test cultures. The
control and test cultures were distributed to a sterile 96-
well plate. Stock solutions of conjugates 21, 22 and 23
were prepared in distilled and deionized water. Then
test solutions at di€erent concentrations were made by
proper dilution of the stock solution. Preformed Fe(III)
complexes of 21, 22 and 23 were made by addition of an
equal volume of test stock solutions and aqueous FeCl3
solution and from those, other test solutions were made
by proper dilution with DI water. Solutions of the test
compounds were added to the 96-well microplate, which
contained sterile Luria broth or sterile Luria broth with
EDDA (100 mg/mL), to give the ®nal concentration of
the conjugates, respectively.
16. Lee, B. H.; Miller, M. J.; Prody, C. A.; Neilands, J. B. J.
Med. Chem. 1985, 28, 323.
17. Ghosh, M.; Miller, M. J. J. Org. Chem. 1994, 59, 1020.
18. Ghosh, M.; Miller, M. J. Bioorg. Med. Chem. 1995, 3,
1519.
19. Ghosh, M.; Miller, M. J. J. Bioorg. Med. Chem. 1996, 4,
43.
20. Franklin, T. J.; Snow, G. A. Biochemistry of Antimicrobial
Action. Chapman and Hall: London, 1988.
21. Just, G.; Grozinger, K. Synthesis 1976, 457.
22. Windholz, T. B.; Johnson, D. B. Tetrahedron Lett. 1967,
27, 2555.

3038
Y. Lu, M. J. Miller / Bioorg. Med. Chem. 7 (1999) 3025±3038
23. Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397.
24. Dong, Q.; Anderson, E.; Ciufolini, M. A. Tetrahedron
Lett. 1995, 35, 5681.
25. Brosius, A. D.; Overman, L. E. J. Org. Chem. 1997, 62, 440.
26. Rogers, H. J. Infect. Immun. 1973, 7, 445.
27. Ong, S. A.; Peterson, T.; Neilands, J. B. J. Biol. Chem.
1979, 254, 1860.
28. Miller, J. H. In Experiments in Molecular Genetics; Cold
Spring Harbor Laboratories: Cold Spring Harbor, 1972; p.
431.