Total synthesis and antifungal evaluation of cyclic aminohexapeptides

Article abstract of DOI:10.1016/S0968-0896(00)00097-3

The need for new therapies to treat systemic fungal infections continues to rise. Naturally occurring hexapeptide echinocandin B (1) has shown potent antifungal activity via its inhibition of the synthesis of β-1,3 glucan, a key fungal cell wall component. Although this series of agents has been limited thus far based on their physicochemical characteristics, we have found that the synthesis of analogues bearing an aminoproline residue in the 'northwest' position imparts greatly improved water solubility (>5 mg/mL). The synthesis and structure-activity relationships (SAR) based on whole cell and upon in vivo activity of the series of compounds are reported. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(00)00097-3

Bioorganic & Medicinal Chemistry 8 (2000) 1677±1696
Total Synthesis and Antifungal Evaluation of Cyclic
Aminohexapeptides
Larry L. Klein,a,* Leping Li,b Hui-Ju Chen,a Cynthia B. Curty,a
David A. DeGoey,a David J. Grampovnik,a Christina L. Leone,a Sheela A. Thomas,a
Clinton M. Yeung,a Kenneth W. Funk,c Vimal Kishore,c Edwin O. Lundell,c
Dariusz Wodka,c Jon A. Meulbroek,a Je€rey D. Alder,d Angela M. Nilius,a
Paul A. Larteye and Jacob J. Plattnerf
aInfectious Disease Research, D47D, AP52, Abbott Laboratories, Abbott Park, IL 60064±3500, USA
bTularik Inc., Two Corporate Drive, South San Francisco, CA 94080, USA
cChemical and Agricultural Division, Abbott Laboratories, North Chicago, IL, USA
dScriptgen Pharmaceuticals, 610 Lincoln Street, Waltham, MA 02451, USA
ePharmacia & Upjohn, 7000 Portage Road, Kalamazoo, MI 49001, USA
fChiron Corporation, 4560 Horton Street, Emeryville, CA 94608, USA
Received 10 December 1999; accepted 9 March 2000
AbstractÐThe need for new therapies to treat systemic fungal infections continues to rise. Naturally occurring hexapeptide echi-
nocandin B (1) has shown potent antifungal activity via its inhibition of the synthesis of b-1,3 glucan, a key fungal cell wall com-
ponent. Although this series of agents has been limited thus far based on their physicochemical characteristics, we have found that
the synthesis of analogues bearing an aminoproline residue in the `northwest' position imparts greatly improved water solubility
(>5 mg/mL). The synthesis and structure±activity relationships (SAR) based on whole cell and upon in vivo activity of the series of
compounds are reported. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
is paramount for successful therapy in patients whose
immune system is challenged.
The recent demand for more e€ective treatment of sys-
temic fungal infections has been due, in part, to the
increase in the immunosuppressed patient population.
The vulnerability of these patients toward commensal
infections caused by organisms such as Candida albicans
and other previously innocuous and rare fungal organ-
isms such as Fusarium has also increased this need.1 The
toxicity of older agents such as amphotericin B2 and the
potential for resistance to the well-known azole agents,
such as ¯uconazole,3 have encouraged further study for
the identi®cation of even more potent analogues or new
agents having unrelated and novel modes of action.
Recently, research e€orts have been focused on the
inhibition of those particular targets which exhibit fun-
gicidal rather than fungistatic properties, since the
penultimate clearance of these slowly invasive infections
b-1,3-Glucan is present as the major structural carbo-
hydrate component in the cell wall of many pathogenic
fungi.4 The producing enzyme, b-1,3-glucan synthetase,
is a membrane-bound protein utilizing UDP-glucose as
the substrate for polymer synthesis. Along with other
components, including a GDP-binding protein,5a this
complex extrudes glucan micro®brils into the interstitial
space of the cell where other glycosyltransferases com-
plete the crosslinking steps required to construct the
growing cell wall. It has been shown that interference of
this process leads to a weakened cell wall and sub-
sequent cell content leakage due to the internal osmotic
pressure ultimately results in cell death.6
Over the last 10 years the search for novel inhibitors of
glucan synthesis has primarily identi®ed two naturally
occurring structural classes which have entered clinical
study or have shown encouraging preclinical data: (1)
the lipopeptide series, typi®ed by echinocandin B (1)7
*Corresponding author. Tel.: +1-847-937-4086; fax: +1-847-938-
2756; e-mail: larry.l.klein@abbott.com
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00097-3

1678
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
(Fig. 1) and (2) the liposaccharide series, typi®ed by
papulacandin.8 Both series of compounds exhibit potent
inhibition of glucan synthesis and display whole cell
antifungal activity; however, only the lipopeptide series
has exhibited good in vivo activity in rodent models
which was similar to that of the industry standard,
amphotericin B. A lead series of modi®ed echinocandin
B analogues was prepared by the Lilly group leading
to LY-303366 (2)9 as a potential drug candidate. The
Merck group isolated pneumocandin B0 (3a) and,
following several functional group modi®cations, pro-
duced L-743,872 (3b) as a clinical candidate.10
have found that this aminoproline (Amp) moiety
improves both the water solubility and antifungal
activity relative to those of the proline or hydroxypro-
line fragments. Preliminary data toward this end have
been presented from a parallel synthesis study.12 We
report herein the e€ects on antifungal activity of further
modi®cations to several of the amino acid residues in
5(A, B, C) and the lipophilic side-chain in 5(G).
Results and Discussion
Chemistry
In 1992 the total synthesis of simpli®ed echinocandin
analogues using solid-phase techniques was reported by
Merck.11 Structure±activity relationship data from this
work showed that several of the functional groups, pri-
marily the hydroxyl groups, were not necessary for
antifungal activity, leading to simpli®ed analogues
which retained reasonable antifungal activity such as 4
(Fig. 2; compare to structure 1). This result encouraged
us to investigate extending this approach in a more
comprehensive manner toward further simpli®cation of
this hexapeptide agent. Furthermore, in order to
increase the solubility of these defunctionalized analo-
gues, we replaced the 4-methyl-3-hydroxyproline residue
of the echinocandins with cis-4-amino-(l)-proline. We
The initial stages of this work utilized solid phase
synthesis for preparation of the acyclic amino acids. The
homotyrosine (hTyr) derivative 6b was protected as the
trimethylsilylethyl (TSE) ester 7 (Scheme 1). This resi-
due was attached to p-carboxymethoxybenzyl linker via
the phenolic hydroxyl group to give 8b and loaded onto
the BHA resin (see Experimental). Removal of the
¯uorenylmethoxycarbonyl (Fmoc) group produced 9b
which has the resin-linked hTyr available for chain
elongation. In order to procure larger quantities of
products for biological testing, preparation of analogues
was also carried out via a solution phase approach
starting with the di-t-butyl protected hTyr derivative 9a.
Figure 1.
Figure 2.

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
1679
Scheme 1.
Free hydroxyl groups and the C-terminal acids of other
residues were generally protected as the corresponding
t-butyl ethers and esters as needed with the N-terminus
protected as its Fmoc derivative. Preparation of the
azidoproline building blocks used in all approaches
commenced from the commercially available BocHyp
OMe (10a) (Scheme 2) via the trans-4-mesylate followed
by azide displacement. The product azide was either
unblocked to amine 12, or hydrolyzed and reprotected
as Fmoc-acid 13. A speci®c example of the hexapeptide
preparation is shown in Schemes 3±5. The majority of
changes to the proline and threonine positions (residues
A and B in 5) were carried out via the intermediate tetra-
peptides 14a and 14b (Scheme 3), which were, in turn,
prepared via a linear coupling/deprotection sequence
starting from the deprotected or resin bound hTyr 9a or
9b, respectively.13 The aromatic side-chains were cou-
pled by initial removal of the Fmoc group of 14 and
subsequent acylation of this amine via the active side-
chain esters as shown for 15, typically as the corre-
sponding 2,4,5-trichlorophenyl esters.9 Hydrogenolytic
cleavage of the carbobenzyloxycarbonyl (Cbz) group
gave the free amine aminopeptide 15. Standard solution
phase coupling conditions were used to extend the chain
to hexapeptide 17 in a linear sequence using 13 followed
by BocThr(Ot-Bu)OH. Precyclitive deprotection of the
t-butyl ethers, carbamates, and esters was carried out using
90% tri¯uoroacetic acid (TFA) at ambient temperature to
give the unprotected hexapeptide 17. Direct ring closure of
this hexapeptide was carried out using either diphenylpho-
sphoryl azide (DPPA) or benzotriazole-1-yl-oxy-trispyrro-
lidinophosphonium hexa¯uorophosphate (PyBOP) and
usually a€orded 30±60% yield of the cyclized product.
Reduction of the azido group with triphenylphosphine
(TPP) or catalytic hydrogenation with simultaneous
removal of Cbz groups was performed to produce the ®nal
aminopeptide analogues shown in Tables 1 and 3.
Scheme 4 illustrates the modi®ed approach for studying
variations in the hTyr structure. The precursor penta-
peptide, 22, was constructed from the proline residue at
its C-terminus and was coupled with the modi®ed ®nal
amino acid 26 via standard conditions. Furthermore,
the ®nal deprotection, cyclization, and reduction of the
azide in 27 proceeded as for 17 (Scheme 3) to give the
hTyr analogues 28 (Table 4).
Scheme 2.

1680
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
Scheme 3.
Alternatively, the cyclic hexapeptide was constructed
via a `3+3' format as in Scheme 5 where the two tri-
peptides, Azp-AA-Hty 29 and Orn-Thr-AA 30, are ®rst
coupled at the C-D position to form linear hexapeptide
31. Deprotection and cyclization then produces inter-
mediate 32 which has two, or in some cases, three
orthogonally protected amino groups available for fur-
ther modi®cation. Deprotection of the BocOrn group
allows for attachment of the side-chain (Table 5).
Reduction of the azide in 33 with TPP then allows for
alkylation or acylation of the proline amine (Table 2).
Finally, in compounds with the Orn(Z) B residue,
hydrogenolysis is carried out and leads to the ®nal
compounds in Table 6.
These two analogues did show >92% purity from
HPLC analysis and the desired mass spectral data.
Structure±Activity Relationships
Assay descriptions
The mode of action of these compounds has been
determined to be inhibition of b-1,3-glucan synthesis as
previously described.5b The in vitro antifungal activity
was evaluated by broth microdilution using amphoter-
icin B as a control. The in vitro activity was quantitated
visually as the lowest concentration of compound
resulting in inhibition of yeast growth.14 Those analo-
gues which showed promising in vitro activity or bore
novel structural elements were tested in our in vivo
mouse model. This acute model has been described
previously15 and a€orded protection results after 10
Characterization of the ®nal compounds is shown in
Table 7. In addition to mass spectral analysis, high
resolution mass spectral data were obtained for all but
two of the analogues which proved dicult to analyze.

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
1681
Scheme 4.
days. The in vivo data in Tables 1±6 are reported as a
ratio to standard amphotericin B to normalize run-to-
run variation.
this compound was quite insoluble and dicult to
formulate for in vivo testing. Introduction of an amino
group on the proline ring not only increased the water
solubility of these compounds, but also enhanced the in
vitro and in vivo antifungal activities of these analogues.
Structure±activity studies showed that the 4b-amino-
proline analogue 18 was the most active among the
stero- and regioisomeric proline amines listed in Table
1. In addition, ease of synthesis proved to be an advan-
tage for 18 as compared to the 3-aminoproline analogue
19a. While hydroxyprolines 19b and 19c exhibited rea-
sonable in vitro activity, no enhancement of the in vivo
activity was observed. Similarly, aminoproline analogue
19f exhibited a moderate boost only in the in vitro
A. Aminoproline position
Initial structure±activity studies showed that the 4b-
aminoproline analogue, 18, was most active among the
stereo- and regioisomeric proline amines listed in Table
1. Furthermore, 18 also showed improved activity as
compared to the corresponding 4b-hydroxyl analogue
19b. The aminomethylproline16 analogue, 19f, also
exhibited moderate activity. While the unsubstituted
proline analogue 19d exhibited good in vitro activity,

1682
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
Scheme 5.
activity. The aminoproline position proved to be quite
sensitive to gross structural modi®cation of this residue.
For example, increasing the size of the ring as in 19e led
to a relatively inactive compound. Furthermore, we
know from previous work that analogues bearing the
opposite (d)-proline stereochemistry exhibited greatly
decreased activity.17 These results suggest that the pro-
line ring may play an important role in the conforma-
tional character of the macro-ring structure.
naturally occurring pneumocandin compounds contain
a hydroxyglutamine at this position which had been
reduced to an ornithine residue.19 In our system, minor
changes to the threonine group had little e€ect on the
activity (Table 3). Substitution of this residue with ser-
ine 21a, ethylglycine 21b, and even glycine 21c did not
greatly diminish this activity; however, inversion of the
stereochemistry 21d or disubstitution at the a-carbon
21e led to analogues with greatly diminished activity. It
is not clear whether these e€ects are due to a con-
formational change in the ring caused by the a-carbon
con®guration or result from the positioning of the
amino acid side-chain itself. Because of this e€ect, the
(l)-stereochemistry was retained in all other analogues.
The e€ect of aminoproline N-alkyl and N-acyl sub-
stituents on the antifungal activity was also studied
(Table 2). It was found that as the length of the alkyl
residue was extended from methyl 20b to benzyl 20a and
then to octyl 20c, the whole cell antifungal activity
decreased. This same trend was seen in the case of the
acylated amines 20d±g. Other carboxyl derivatives such
as carbamates and ureas, 20k and 20j, were also studied.
In general the bulk of the corresponding alkyl residues
in these series seemed to have a greater e€ect on dimin-
ishing activity than the functionality of these groups.
One of the most active substituents on the proline was
found to be the guanidino group 20h which, in addition
to retaining good activity, also exhibited excellent solu-
bility characteristics.
A wide variety of functional groups at this position were
found to be compatible with retention of good whole
cell antifungal activity. Amine-containing residues (Orn
or Dap) had a distinct advantage over carboxylic acid
groups (Glu) in the in vivo assay. Carboxylic acid
groups arising from glutamic acid 21g, phenols such as
tyrosine 21f, and amines such as ornithine 21h or argi-
nine 21j, all were quite active. In summary, ornithine
was considered as the optimal group for this position
based upon antifungal activity, water solubility, and
cost.
B. Thr position
C. hTyr
Prior to our work, some data were available for mod-
i®cations at the B threonine position such as replace-
ment by serine and glycine as described by Merck11 and
Lilly18 respectively, in their particular systems. The
Previous work by Merck11 established that homo-
tyrosine (hTyr) showed activity similar to that of the
naturally occurring dihydroxyhomotyrosine whereas the

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
Table 1. Antifungal activity of residue A analogues
1683
In vitro activity (mg/mL)
In vivo activity
R1
H
R2
H
Candida albicans
ATCC 62376
Candida albicans
ATCC 38247
Candida glabrata
ATCC 15545
ED50/ED50AmpB
18
0.2
0.2
0.39
0.2
0.39
0.39
1.56
0.78
0.2
10.6
Ð
19a
19b
19c
0.78
3.12
0.39
H
H
Ð
H
H
>20
19d
19e
3.12
6.25
6.25
>100
12.5
12.5
>20
Ð
19f
H
0.78
0.39
0.78
>20
aTer=-(O)C-Ph-Ph-Ph-O-(CH2)4CH3.
Table 2. Antifungal activity of substituted aminoproline analogues
In vitro activity (mg/mL)
In vivo activity
20
R
Candida albicans
ATCC 62376
Candida albicans
ATCC 38247
Candida glabrata
ATCC 15545
ED50/ED50AmpB
a
b
c
d
e
f
g
h
i
Bn
CH3
Octyl
Ac
0.39
0.2
6.25
0.2
6.25
0.39
50
1.25
0.39
6.25
0.39
Ð
8.4
Ð
>20
Ð
Ð
>20
26
0.39
>100
25
3.12
0.39
>100
1.25
0.78
Octanoyl
Benzoyl
Succinoyl
Guanidinyl
Boc
50
>100
6.25
6.25
0.39
100
6.25
0.78
0.39
12.5
0.39
0.2
Ð
Ð
Ð
j
k
CH3OC(O)
CH3NHC(O)
3.12
0.39
corresponding tyrosine and phenylalanine analogues
were much less active. In our system, the results from
the tyrosine (Tyr) 28a and homophenylalanine (hPhe)
analogues 28d (Table 4) were consistent with those pre-
viously described, i.e., both were less active than the
parent hTyr, with Tyr less active of the two. The three
atom oxygen-bridged analogue 28e prepared from ben-
zylserine shows similar activity to its two-carbon parent;
however, the corresponding inverted stereoisomer 28f
was also prepared and, as in the previous cases, exhibited

1684
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
Table 3. Antifungal activity of residue B analogues
In vitro activity (mg/mL)
In vivo activity
21
B
Candida albicans
ATCC 62376
Candida albicans
ATCC 38247
Candida glabrata
ATCC 15545
ED50/ED50AmpB
a
b
c
d
e
f
g
h
i
(l)-Serine
(l)-Ethylglycine
(l)-Glycine
0.1
0.1
0.2
12.5
50
1.56
0.39
0.2
0.2
0.39
1.56
6.25
25
1.56
1.56
0.39
0.78
0.78
0.1
0.1
4.98
6.49
Ð
Ð
Ð
1.56
3.12
6.25
0.78
0.78
0.2
(d)-Threonine
a,a-Dimethylglycine
(l)-Tyrosine
(l)-Glutamic acid
(l)-Ornithine
Ð
>20
3.01
21.8
11.2
(l)-Diaminopropionic acid (Dap)
(l)-Arginine
1.56
0.39
0.39
0.2
j
Table 4. Antifungal activity of residue C analogues
In vitro activity (mg/mL)
In vivo activity
28
R1
R2
R3
R4
Candida albicans
ATCC 62376
Candida albicans
ATCC 38247
Candida glabrata
ATCC 15545
ED50/ED50AmpB
a
b
c
d
e
f
HO
HOCH2
HOCH2
H
CH2
CH2
Tera
PIP8b
PIP8
Ter
Ter
Ter
OH
H
H
OH
OH
OH
H
50
12.5
1.56
1.56
1.56
>100
3.12
0.2
25
6.25
>100
1.56
Ð
Ð
>20
>20
Ð
Ð
Ð
>20
12.5
3.12
3.12
3.12
25
CH2CH2
CH2CH2
CH2-O-CH2
CH2-O-CH2
CH2CH2
CH2CH2
3.12
3.12
25
3.12
0.78
H
H (D)
CH3
NH2
g
h
PIP8
PIP8
6.25
0.78
H
aTer=-(O)C-Ph-Ph-Ph-O-(CH2)4CH3.
bPIP8=(O)C-Ph-piperazine-Ph-O-(CH2)6CH3.
much lower activity. The importance of the phenolic
hydroxyl group was studied by testing the hydroxy-
methyl 28c, methyl 28g, and amino 28h group replace-
ments, though only the latter derivative had near
optimal antifungal activity. All one-carbon linked ana-
logues 28a,b exhibited poor activity regardless of reten-
tion of the phenolic or hydroxymethyl group. These
SAR results suggest that these chain shortened analo-
gues may su€er a conformational change due to closer
interactions of the aryl group with the macrocycle ring.
D. Side-chain
Problems related to hemolysis led the Lilly group to
modify the side-chain of echinocandin B.9 Their SAR
work led initially to a 4-(n-octyloxy)benzoyl(p-OB)
chain (Cilofungin, 2, R=p-OB), which later evolved to
the 4-(n-pentyloxy)terphenoyl group as being the opti-
mal side-chain for their system. Aspects of lipophilicity,
length, and shape of this side-chain were also considered
as variables in our work, though not as comprehensively

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
Table 5. Antifungal activity of side-chain analogues
1685
In vitro activity (mg/mL)
In vivo activity
34
R
Candida albicans
ATCC 62376
Candida albicans
ATCC 38247
Candida glabrata
ATCC 15545
ED50/ED50AmpB
a
b
c
Ac
C(O)(CH2)13CH3
Bz
>100
0.78
>100
0.2
>100
0.78
>100
0.39
>100
1.25
>100
12.5
Ð
>20
Ð
d
4
e
f
>100
25
>100
50
>100
50
Ð
Ð
g
h
i
j
k
CH3
C5H11
C8H17
C12H25
12.5
0.2
0.2
0.78
0.2
3.12
0.2
0.2
1.56
0.2
12.5
0.69
0.39
0.39
0.2
Ð
>20
7.63
>20
10
CH2-(c-C6H11)
Table 6. Antifungal activity of combination analogues
In vitro activity (mg/mL)
In vivo activity
35
R1
R2
R3
R4
Candida albicans
ATCC 62376
Candida albicans
ATCC 38247
Candida glabrata
ATCC 15545
ED50/ED50AmpB
a
b
c
d
e
f
NH2
Guanidino
NH2
Guanidino
NH2
Guanidino
NH2
Guanidino
NH2
NH2
NH2
OH
OH
OH
OH
OH
OH
H
n-C5H11
n-C5H11
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C8H17
0.78
0.39
0.78
0.39
0.2
0.78
0.2
0.2
0.78
1.56
0.78
0.78
0.39
0.78
0.39
0.2
0.78
1.56
0.78
0.39
0.2
0.39
0.2
0.2
13.6
13.5
3.8
2.16
1.4
1.5
0.38
0.2
NH2
H2NCH2CH2
H2NCH2CH2
H2NCH2CH2
H2NCH2CH2
g
h
H
as in the previous study. Whereas the acetyl and benzoyl
analogues (34a and 34c) are inactive on our hexapeptide
(Table 5), extending or appending a hydrophobic chain
onto either recovers the activity. Side-chain length and
rigidity components were further studied by addition of
ring units, e.g., from 4-(n-octyloxy)benzoyl, to 40-(octyl-
oxyphenyl)benzoyl and 400-(n-octyloxy)terphenoyl, which
showed increasing activity.
The Fujisawa group20 patented a pneumocandin-like
series bearing a l,4-diphenylpiperazine based side-chain
for improved potency which also increased water solu-
bility. We prepared several analogues using this side-
chain and also several truncated versions which showed
that an unsubstituted piperazine 34e or phenolic group
34f led to a loss of activity. As the length of the alkyl
residue (R2 in Table 5) is increased to an eight carbon

1686
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
Table 7. Characterization of target compounds
Compound
MS (FAB)
m/z
HRMS
Calculated for Measured
HPLCa
%/tR
Compound
MS (FAB)
m/z
HRMS
Calculated for Measured
HPLCa
%/tR
Table 7a
18
Table 7d
28a
1061
[M+H+
1099
C57H73N8O12 1061.5363
1061.5348
C57H73N8O12 1061.5327
1061.5348
C57H71N7O13 1062.5165
1062.5188
100/13.9
100/13.4
81/7.4
1047
[M+H+
1095
C56H70N8O12 1047.516
1047.5191
C58H83N10O11 1095.6238
1095.6243
83/14.2e
98/10.7
93/11.3
96/12.3d
98/12.4d
90/12.5
98/13.9
97/10.6
]
]
]
]
]
]
]
]
]
]
]
]
]
]
]
19a
19b
19c
19d
19e
19f
28b
28c
28d
28e
28f
28g
28h
[M+K+
1062
[M+H+
1109
C59H85N10O11
f
[M+H+
1062
[M+H+
1045
C
C
C
57H71N7O13 1062.5182
1062.5188
57H71N7O12 1046.5263
1046.5239
58H74N7O12 1060.5408
1060.5395
100/14.7
95/16.2
87/16.3
100/13.3
C
C
57H72N8O11 1045.540
1045.5399
57H72N8O12 1061.537
1061.5348
58H83N10O11 1095.6245
[M+H+
1046
[M+H+
1061
[M+H+
1060
[M+H+
1095
C
[M+H+
1075
[M+H+
1094
1095.6243
C59H85N10O10 1094.6450
1094.6450
C58H75N8O12 1075.5485
1075.5504
[M+H+
[M+H+
1094
Table 7b
20a
C58H84N11O10
f
1151
[M+H+
1075
C64N79N8O12 1151.5844
1151.5817
83/17.65
100/14.4
84/22
[M+H+
]
]
]
]
]
]
]
]
]
]
]
Table 7e
34a
20b
20c
20d
20e
20f
20g
20h
20i
C58H75N8O12 1075.5514
1075.5504
761
[M+H+
943
C
35H53N8O11
761.3834
761.3853 89.9/21.4
943.5860 92.8/13.6
823.3981 97.1/3.8
[M+H+
1173
]
]
]
]
]
]
]
]
]
]
]
C65H89N8O12 1173.6625
1173.6600
C59H75N8O13 1103.5469
1103.5454
C65H87N8O13 1187.6409
1187.6393
C64H77N8O13 1165.5626
1165.5610
34b
34c
34d
34e
34f
34g
34h
34i
C48H78N8O11
943.5868
C40H54N8O11
823.3990
[M+H+
1103
[M+H+
823
100/14.7
79/12.0
[M+H+
1187
[M+H+
1027
C54H74N8O12 1027.5507
1027.5504
90/14.6
[M+H+
1203
[M+H+
907
100/18.6
100/13.6
100/14.2
90/19.2
C44H62N10O11
907.4678
907.4694 94.3/8.1b
[M+K+
1161
[M+H+
999
C61H77N8O15 1161.5491
1161.5508
C
C
C
50H67N10O12
999.4959 89/3.8
[M+H+
1103
[M+H+
1013
999.4940
51H68N10O12 1013.5095 88.2/3.8
1013.5096
55H76N10O12 1069.5726 93.1/5.5
1069.5722
C
58H75N10O12 1103.5544
1103.5566
[M+H+
1161
[M+H+
1069
C62H81N8O14 1161.5875
1161.5872
[M+H+
1119
[M+H+
1111
20j
C59H75N8O14 1119.541
1119.5403
C59H76N9O13 1118.5575
1118.5563
100/15.8
100/14.6
C58H83N10O12 1111.6188 92.8/10
1111.6192
C62H91N10O12 1167.6830
1167.6818
C57H79N10O12 1095.5903
1095.5879
[M+H+
1156
[M+H+
1167
20k
34j
83/18
[M+K+
[M+H+
1095
Table 7c
21a
34k
93/7
1085
[M+K+
1083
C
56H70N8O12
1085.4750
K
K
K
K
1085.4745
1083.4937
1055.4653
1085.4745
98/13.2
99/13.8
97/13.2
91/13.2
98/14.1
99/14.1c
100/13.2
100/11.7
98/11.0c
95/12.1
[M+H+
]
]
]
]
]
]
]
]
]
]
Table 7f
35a
21b
21c
21d
21e
21f
21g
21h
21i
C57H72N8O11
1083.4958
C55H68N8O11
1055.4645
C56H70N8O12
1085.4750
1054
[M+H+
1096
C54H76N11O11 1054.5740
1054.5726
C55H78N13O11 1096.5917
1096.5944
C57H82N11O10 1095.6201
1095.615
C58H84N13O11 1138.6416
1138.6413
70/5.4
87/5.1
90/9.8
90/9.8
95/8.9
90/9.3
98/11.8e
98/11.9e
[M+K+
1055
]
]
]
]
]
]
]
]
35b
35c
35d
35e
35f
35g
35h
[M+K+
1085
[M+H+
1096
[M+K+
1045
[M+H+
1138
C57H73N8O11 1045.5399
1045.5399
C62H75N8O12 1123.547
1123.5504
[M+H+
1123
[M+H+
1124
C59H86N11O11 1124.6493
1124.6508
[M+H+
1089
[M+H+
1166
C58H73N8O13 1089.5293
C
C
C
60H88N13O11 1166.6686
1166.6726
59H86N11O10 1108.6534
1108.6559
60H87N13O10 1150.677
1150.6777
[M+H+
1112
1089.5297
58H75N9O11
[M+H+
1108
C
K
1112.5223
[M+K+
1084
1112.5223
56H72N9O11 1046.5347
[M+H+
1150
C
[M+K+
1116
1046.5351
59H78N11O11 1116.588
1116.5882
[M+H+
21j
C
[M+H+
aPurity determined by reversed phase HPLC analysis. Unless otherwise stated, reversed phase HPLC analysis was carried out using the following
method: column: dynamax C18 (25 cm L, 4.6 mm ID, 5 m, 300 A); ¯ow rate: 1 mL/min; mobile phase: B acetonitrile, A 0.1% TFA in water;
gradient: linear from 45 to 75% B in 14.56 min, from 75 to 95% B in 0.44 min; detection: UV, 254 nm.
bSame method as (a) except: gradient: linear from 10 to 75% B in 14.56 min.
cSame method as (a) except: ¯ow rate: 2 mL/min; column: Waters Delta-Pak C18 (20 cm L, 8 mm ID, 15 m, 300 A).
dSame method as (c) except: gradient: linear from 45 to 95% B in 15 min.
eSame method as (c) except: gradient: linear from 25 to 95% B in 15 min.
fWe were unable to obtain HRMS data of these samples prior to exhaustion of sample although the mass spectral and HPLC data clearly showed
the expected molecular weight and purity.

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
1687
chain as for 34i, antifungal activity increases, whereas
longer residues show poorer results. In terms of the whole
cell and in vivo antifungal activity, and water solubility,
we found the octyloxyphenyl-piperazinylbenzoyl 34i and
cyclohexylmethyl 34k side-chains to be optimal.
and brine, and then dried over Na2SO4, ®ltered and the
solvent evaporated. All puri®cations were performed
with silica gel (E. Merck silica gel 60, 230±400 mesh)
¯ash chromatography unless otherwise stated. Methyl-
ene chloride (DCM) was distilled from calcium hydride,
and tetrahydrofuran (THF) was distilled from sodium±
benzophenone. The reagents used for coupling were N-
hydroxybenzotriazole (HOBT), 1-(3-dimethylamino-
propyl)-3-ethylcarbodiimide hydrochloride (EDAC),
diisopropylcarbodiimide (DIPC), dicyclohexylcarbo-
diimide (DCC), N-ethoxycarbonyl-2-ethoxy-1,2-de-
hydroquinoline (EEDQ, 1-cyclohexyl-3-(2-morphol-
inoethyl)carbodiimide metho-p-toluenesulfonate (CMC),
and benzotriazole-1-yl-oxy-tris-pyrrolidinophosphonium
hexa¯uorophosphate (PyBOP). Mass spectral data were
recorded with a Finnigan SSQ 7000 instrument, and
high resolution MS were obtained on a Kratos MS 50
instrument.
E. Other residues
Regarding the remaining residues in the cyclic hexapep-
tide, it was shown that replacing the Hyp D (Fig. 2) ring
with Amp, similar to the A `corner' of the molecule, led
to compounds showing poor activity; however, substitu-
tion of Hyp with Pro a€orded compounds which
retained activity but minimized cost and protecting
group requirements.21 Other modi®cations to the D, E,
and F residues that have been carried out will be reported
at a later date.22
F. Combination analogues
Biological evaluation
The SAR information from the previous studies was
used to design and prepare an optimized set of analo-
gues relative to their potency, water solubility,23 and
ecacy. The optimum groups were chosen from the
study of residue A (4b-amino and guanidino groups),
from residue B (diaminopropionic (Dap) acid and orni-
thine (Orn) groups), and from residue G (n-octyl- and n-
pentyloxyphenylpiperazinylbenzoyl side-chains) (Table
6). The guanidinoproline group showed equal or slightly
better activity as compared to the Amp group with
improved solubility characteristics, and thus was con-
sidered the best choice at residue A. At residue B, pla-
cement of Dap resulted in analogues 35c,d with no clear
advantage over the less costly Orn analogues 35e,f.
Optimum ecacy was obtained with the octyloxy moi-
ety (35c,d) in the side-chain residue G over the pentyl-
oxy analogues 35a,b. While the last four entries in Table
6 were all quite potent as antifungal agents both in the
whole cell and mouse model assays, entry 35h was
chosen as the best compound in our series based on the
total biological and chemical pro®le.
Whole cell antifungal assay. Antifungal activity was
evaluated by broth microdilution. Compounds were
dissolved in DMSO, then diluted in yeast nitrogen broth
(Difco Laboratories, Detroit, MI) containing 5 g/L
dextrose, pH 7.0, in serial 2-fold steps. Amphotericin B
was included as a control. The tests were inoculated
with the yeast strains at approximately 104 colony-
forming units per mL prepared from an overnight cul-
ture on Sabouraud dextrose agar. The cultures were
diluted to a turbidity equivalent to 0.5 McFarland
standard, then inoculated at a ®nal dilution of 1:100.
Tests were incubated for 18±20 h at 35 ꢀC. The in vitro
activity was determined visually as the lowest con-
centration of compound resulting in inhibition of yeast
growth. The yeast strains were clinical isolates pur-
chased from the American Type Culture Collection,
Rockville, MD. C. albicans ATCC 62376 and C. glab-
rata ATCC 15545 were susceptible isolates and C. albi-
cans ATCC 38247 is resistant to polyenes.
Acute Candida mouse model.24 For ecacy evaluation
of antifungal compounds, female CF-I outbred mice
(Charles River Laboratories) of approximately 8 weeks
of age and 25 g body mass were inoculated intrave-
nously with undiluted overnight culture of C. albicans
CAF2 grown in Sabauroud broth. The inocula were
approximately 2Â106 cfu/mouse, and generally resulted
in mortality of untreated mice within 24 h post inocu-
lation. Test compounds were formulated in 2% DMSO/
PBS for injection and were administered by intra-
peritoneal injection at 1 h post inoculation. There were
10 mice per dosage group.
Conclusion
We have shown that incorporation of an aminoproline
residue into the ring of these cyclic hexapeptide anti-
fungal agents leads to improvements of potency, in vivo
ecacy, along with obvious e€ects on physico-chemical
characteristics such as water solubility. We have shown
that by combining several optimally modi®ed moieties
in this series, the resultant analogue exhibits 5-fold
greater in vivo activity in our mouse model as compared
to amphotericin B.
Mice were monitored daily for survival for 10 days. The
percent survival rate indicates the percentage of mice
per treatment group surviving to the termination of the
trial at day 10 post inoculation. Curative therapy was
determined by plating kidney homogenates of individual
surviving mice on Sabouraud dextose agar. Kidney
homogenates absent of C. albicans growth were indicative
of curative ecacy. The percent cure rate indicates the
percentage of mice per treatment group with sterilized
Experimental
All reactions were performed at room temperature
(25 ꢀC) unless otherwise stated. Standard work up refers
to quenching coupling mixtures with methylene chloride
(DCM) and 10% citric acid. The organic layer was
washed consecutively with saturated NaHCO3, H2O,

1688
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
kidneys. The mean e€ective dose to protect 50% of the
mice (ED50 value) was calculated from cumulative
mortality by logarithmic-probit analysis of a plotted
curve of survival versus dosage. The mean ED50 value
for the control, amphotericin B, was 1.62 mg/kg with
the standard deviation (S.D.) being 1.148.
added 2 N sodium hydroxide solution (100 mL) and the
mixture was stirred for 1 h. After evaporation of the
methanol, dioxane (275 mL) was added along with
concd HCl to pH 2 followed by more HCl (19 mL). The
mixture was warmed to 50 ꢀC for 2 h and cooled to 0 ꢀC.
Sodium carbonate (32.3 g, 0.3 mol) was added to pH 9
followed by Fmoc-OSu (27 g, 80 mmol) in dioxane (75
mL). After 2 h at 0 ꢀC, the mixture was warmed to 25 ꢀC
for 18 h. The mixture was ®ltered through Celite, dilu-
ted with ethyl acetate (1.2 L) and washed with 2 N HCl
(250 mL), brine, dried, and the solvent was evaporated.
The residue was puri®ed using 1% methanol/DCM with
0.1% acetic acid to give 22.8 g (60%) of 13.
Fmoc-hTyrOH (6b). To a 0 ꢀC solution of homotyrosine
hydrobromide 6a (4 g, 15 mmol) in 30% aq dioxane (40
mL) was added 20% aq sodium carbonate (35 mL) fol-
lowed by a solution of FmocCl (4 g, 16 mmol) in diox-
ane (20 mL). The mixture was allowed to warm to 25 ꢀC
for 2 h and neutralized with 2 N HCl. The mixture was
extracted twice with ethyl acetate, the organic layer
washed with brine, dried over sodium sulfate, and the
solvents were evaporated. The crude residue was pur-
i®ed by chromatography using hexane:ethyl acetate
(2:1) to give 4.67 g (77%) of 6b.
Tetrapeptide Fmoc(Z)Orn-Thr(Ot-Bu)-Hyp(Ot-Bu)-
hTyr(Ot-Bu)Ot-Bu (14a). (i) To a 0 ꢀC solution of 9a
(8.6 g, 28 mmol) and FmocHyp(OtBu) (11.43 g, 28
mmol) were added CMC (17.74 g, 42 mmol) and HOBT
(5.66 g, 42 mmol), and the mixture was stirred overnight
and allowed to warm to 25 ꢀC. After the standard work
up, the residue was puri®ed using 2±15% ethyl acetate/
DCM as the eluent to give 16.7 g (86%) of FmocHyp-
(Ot-Bu)hTyr(Ot-Bu)Ot-Bu.
hTyr(Ot-Bu)Ot-Bu (9a). A mixture of 6b (20.8 g, 50
mmol) and isobutylene (400 mL) was combined in a
metal bomb with sulfuric acid (4.2 mL) and DCM (400
mL) for 3 days. Saturated sodium bicarbonate was
added (400 mL) and the aqueous layer extracted twice
with DCM. The solvents were evaporated and the crude
residue was puri®ed using hexane:ethyl acetate (8:1±4:1)
to give 16.7 g (63%) of the di-t-butyl Fmoc-hTyr. This
material was dissolved in DMF (250 mL) and treated
with diethylamine (30 mL, 10 equiv) for 4 h. The sol-
vents were evaporated to give crude 9a which was used
directly for the synthesis of tetrapeptide 14a.
(ii) The product from above was treated with diethyl-
amine (25 mL) in DMF (275 mL) for 2 h. The solvents
were evaporated, and the crude residue was dissolved in
DCM (200 mL) and cooled to 0 ꢀC. To this solution
were added FmocThr(Ot-Bu) (10.3 g, 24 mmol), CMC
(15.2 g, 36 mmol), and HOBT (4.8 g, 36 mmol) and the
mixture was stirred overnight and allowed to warm to
25 ꢀC. After the standard work up, the residue was pur-
i®ed with 2.5% ethyl acetate/DCM to give 17.3 g (84%)
of FmocThr(Ot-Bu)-Hyp(Ot-Bu)-hTyr(Ot-Bu)Ot-Bu.
N-Boc-4-(R)-mesyl-(L)-proline methyl ester (10b). A
solution of Boc-(R)-HypOMe (10a) (6.2 g, 25.3 mmol)
in pyridine (50 mL) was treated with methanesulfonyl
chloride (4 mL) at 0 ꢀC, and was warmed to 25 ꢀC
overnight. The reaction mixture was diluted with ethyl
acetate and washed with 0.5 N HCl and brine. The
organic layer was separated and the solvent was evapo-
rated. The residue was triturated with heptane and the
solids were ®ltered to give 8 g (98%) of 10b.
(iii) Preparation of FmocOrn(Z)OH: To a 0 ꢀC suspen-
sion of Orn(Z)OH (19.9 g, 75 mmol) in dioxane (190
mL) and aqueous sodium carbonate (9 g, 84 mmol in 70
mL water) was added FmocOSu (27.8 g, 82 mmol). The
mixture was stirred overnight and allowed to warm to
25 ꢀC. The mixture was diluted with ethyl acetate and
water, and the pH was adjusted to 1 before twice
extracting with ethyl acetate, drying over Na2SO4, and
evaporating the solvent to give FmocOrn(Z)OH which
was used directly in the next step.
N-Boc-4-(S)-azido-(L)-proline methyl ester (11). A solu-
tion of 10b (7.5 g, 23 mmol) and sodium azide (7.5 g,
115 mmol) in dimethylformamide (30 mL) was heated at
60 ꢀC overnight. After cooling to 25 ꢀC the reaction
mixture was diluted with ethyl acetate and washed with
water. The aqueous layer was extracted with ethyl ace-
tate, the organic layers combined, and washed with
brine. The solvents were evaporated to give 6.74 g of
BocAzpOMe (11) as an oil which was used directly in
the next step.
(iv) The product from (ii) above (15.2 g, 24 mmol) was
treated with diethylamine (19 mL) in DMF (230 mL)
for 2 h. The solvents were evaporated and the crude
residue was dissolved in DCM (150 mL) and cooled to
0 ꢀC. To this solution was added FmocOrn(Z)OH (8.75
g, 18 mmol) from (iii) above, CMC (11.3 g, 27 mmol)
and HOBT (3.63 g, 27 mmol) and the mixture was stir-
red overnight and allowed to warm to 25 ꢀC. After the
standard work up the residue was puri®ed using 25±
50% ethyl acetate/DCM to give 16.8 g (67%) of
Fmoc(Z)Orn-Thr(Ot-Bu)-Hyp(Ot-Bu)-Tyr(Ot-Bu)Ot-Bu
14a.
4-(S)-Azido-(L)-proline methyl ester hydrochloride (12).
A solution of 11 in 2.8 N HCl in ethyl acetate (260
mL) was stirred for 1 h. The solvent was evaporated and
the residue treated with methyl t-butyl ether (MTBE) at
0 ꢀC. The solids were ®ltered, washed with MTBE, and
dried in vacuo for 2 h to give 12 which was used directly
for the next step.
RCONHOrn-Thr(Ot-Bu)-Hyp(Ot-Bu)-hTyr(Ot-Bu)Ot-Bu
(15a). (i) Tetrapeptide 14a (0.5 g, 0.52 mmol) was
treated with diethylamine (0.5 mL) in DMF (25 mL) for
2 h. The solvents were evaporated and the residue was
Fmoc-4-(S)-AzpOH (13). To a solution of 12 (27.3 g,
0.1 mol) in methanol (50 mL) and dioxane (50 mL) was

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
1689
treated with 2,4,5-trichlorophenyl 4-(40-pentyloxybi-
pheny1)benzoate20 (0.31 g, 0.57 mmol) in DMF (60 mL)
at 60 ꢀC for 3 h. The solvents were evaporated, and the
residue was puri®ed using 50% ethyl acetate/hexane as
the eluent to give 0.45 g (69%) of the acylated product.
BocThr(O-t-Bu)OH gave 69% 17. Deprotection with
TFA and cyclization a€orded 24% azido intermediate
which was reduced with 10% Pd/C in ethanol to give
19a in 41% yield after HPLC puri®cation.
19b. Coupling to resin 15b (0.52 g, 0.5 mmol/g) with
Fmoc-3b-HypOH was followed by Fmoc deprotection,
coupling to BocThrOH, and cleavage to the acyclic
hexapeptide in 28% total yield. Cyclization a€orded
15% of 19b.
(ii) The product from above (0.5 g, 0.4 mmol) was trea-
ted with Pd(OH)2 on carbon (20% Pd by wt, 83 mg) in
ethanol (10 mL) and acetic acid (0.065 mL). A hydrogen
balloon was used for the reduction which took place
over 4 h. The catalyst was ®ltered and the solvents were
evaporated. The residue was dissolved in ethyl acetate
and washed consecutively with saturated sodium bicar-
bonate, water, and brine and dried over sodium sulfate.
The solvents were evaporated and the product amine
15a (74 mg, 100%) was used directly for the next step.
19c. Coupling to 15a in a similar manner with Fmoc-3a-
HypOH, 87%; deprotection/coupling to BocThr(O-t-
Bu)OH, 80%; TFA deprotection/cyclization to 19c,
43%.
19d. Coupling to resin 15b (1 g, 0.46 mmol/g) with
FmocProOH (see section B.a below) was followed by
Fmoc deprotection and coupling to BocThrOH, and
cleavage to acyclic hexapeptide amino acid, 47%.
Cyclization as above gave 45% of 19d.
Azido pentapeptide (16a). To a solution of 15a (5 g, 4.6
mmol) in DCM (50 mL) at 0 ꢀC was added 13 (2.3 g, 6
mmol), CMC (3 g, 7.1 mmol) and HOBT (1 g, 7.4
mmol) and the mixture was stirred overnight at 25 ꢀC.
After the standard work up, the residue was puri®ed
using ethyl acetate:hexane (7:3) to give 5.5 g (83%) of
the protected pentapeptide 16a.
19e. Coupling to resin 15b (0.5 g, 0.5 mmol/g) with
Fmoc hProOH was followed by Fmoc deprotection,
coupling to BocThrOH, and cleavage to acyclic hexa-
peptide, 58%. Cyclization as above gave 89% of 19e.
Azido hexapeptide (17a). Pentapeptide 16a (2.29 g, 1.5
mmol) was treated with diethylamine (1.7 mL) in DMF
(20 mL) for 2 h. The solvents were evaporated and the
residue was dissolved in DCM (25 mL) at 0 ꢀC. To this
solution was added the N-Boc amino acid (2.1 equiv)
(for 18: BocThr(t-BuO)OH, 0.85 g, 3.l mmol), CMC
(1.99 g, 4.7 mmol) and HOBT (0.63 g, 4.7 mmol) and
the mixture was warmed to 25 ꢀC overnight. Standard
work up a€orded 1.99 g (85%) crude 17.
19f. Coupling to 15a with Fmoc-4-carbobenzyloxy-
aminomethyl-ProOH, 97%. In this case the Fmoc
deprotection/coupling steps and the TFA deprotection/
cyclization/hydrogenolytic deprotection were carried
out directly without puri®cation to give 19f (31% total
yield for ®ve steps).
b. Preparation of N-alkylated prolines (20a±c). 20a. To
a solution of 18 (0.2 g, 0.19 mmol) in dichloroethane (10
mL) and DMF (10 mL) was added benzaldehyde (0.022
mL, 0.21 mmol) followed by sodium cyanoborohydride
(0.016 g) and acetic acid (0.011 mL) and the reaction
mixture was stirred for 18 h. The solvents were evapo-
rated and the residue was dissolved in methanol:DCM
(8 mL, 3:1) and precipitated with water (35 mL). The
crude solids were puri®ed using 5% methanol/DCM
with 0.5% ammonium hydroxide to give 0.13 g (66%)
20a.
Cyclic hexapeptide (18). Product 17 (1.08 mmol) was
treated with 90% tri¯uoroacetic acid (10 mL) at 0±25 ꢀC
for 5 h. The solvent was evaporated and the crude resi-
due was treated with PyBOP (1.68 g, 3.2 mmol), HOBT
(0.44 g, 3.2 mmol) and diisopropylethylamine (DIPEA)
(1.23 mL) in DMF (1000 mL) for 24 h. The solvent was
evaporated and the residue was dissolved with minimal
amounts of methanol followed by water to e€ect pre-
cipitation. The solid product was ®ltered and puri®ed
using 5±12.5% methanol in DCM to give 0.977 g (68%)
of the cyclized azide. This azide was treated with 10%
Pd/C (0.2 g) in methanol (60 mL) with a hydrogen bal-
loon overnight. The reaction mixture was ®ltered and
the solvent was evaporated. The crude product was
puri®ed using 15±20% methanol/10% ammonia/DCM
to give 0.52 g (66%) of 18.
20b. To a solution of 20a (0.13 g, 0.11 mmol) in DMF (4
mL) were added 37% formaldehyde (0.5 mL), sodium
cyanoborohydride (0.014 g), and acetic acid (0.013 mL)
to give after puri®cation using 10% methanol/DCM/
0.5% ammonium hydroxide 0.1 g (84%) of the N-ben-
zyl-N-methyl intermediate.
Section A. Preparation of residue A analogues (19,20)
This intermediate (0.07 g, 0.06 mmol) was treated with
5% Pd/C (0.07 g) in methanol:chloroform (10 mL, 4:1)
and shaken under hydrogen balloon pressure for 18 h.
The catalyst was ®ltered, the solvents evaporated, and
the residue puri®ed using 20% methanol/DCM/0.5%
ammonium hydroxide to give 0.064 g (100%) of 20b.
a. Preparation of aminoproline replacements (19a±f).
Residue A analogues appearing in Table 1 were pre-
pared in a similar manner from 14a as 18. The amino-
methyl proline leading to analogue 19f was prepared as
in ref 16.
19a. Intermediate 15a was treated with the Fmoc-3b-
azidoProOH as in preparation of 16a above to give 82%
pentapeptide 16. The deprotection and coupling to
20c. In a similar manner to 20a, octanal was used to
alkylate 18 to give after puri®cation using 15% methanol/
DCM/0.5% ammonium hydroxide 0.03 g (36%) 20c.

1690
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
c. Preparation of N-acyl proline analogues (20d±k). 20d.
Compound 18 (0.15 g) was treated with acetic anhy-
dride (0.014 mL) in DMF (4 mL) for 3 h. The solvents
were evaporated and the residue puri®ed using 10%
methanol/DCM/0.5% ammonium hydroxide to give
0.1 g (66%) 20d.
puri®ed using 10% methanol/DCM to give 33 mg
(59%) 20k.
Section B. Solid-phase synthesis of hexapeptides:
preparation of residue B analogues
Fmoc-Homotyroslne trimethylsilylethyl (TSE) ester (7).
To a cooled solution of EDAC (2.35 g, 12.3 mmol), tri-
methylsilylethanol (7.5 mL) and catalytic DMAP in
THF (30 mL) was added 6b (4.4 g, 11 mmol) in THF
(30 mL). After 5 h the reaction was diluted with ethyl
ether and washed with 2 N HCl. The organic layer was
separated, washed with brine and the solvent was eva-
porated. The residue was puri®ed using hexane:ethyl
acetate (5:1) to give 4 g (73%) of the ester 7.
20e. A solution of 2,4,5-trichlorophenyl octanoate was
prepared by combining octanoic acid (9.4 mL, 0.062
mmol) 2,4,5-trichlorophenol (12.3 mg, 0.062 mmol),
DCC (12.8 mg, 0.062 mmol) and catalytic DMAP in
DCM (2 mL) at 0 ꢀC for l h. This mixture was ®ltered
and combined with 18 (30 mg, 0.028 mmol) and DIPEA
(5 mL) in DMF (0.5 mL). This mixture was stirred
overnight, the solvents were evaporated, and the crude
residue was puri®ed using 10% MeOH/DCM to give 16
mg (50%) 20e.
Fmoc-Homotyrosine 40-allyloxyphenylmethyl ether TSE
ester (8a). To a cooled solution of 7 (4 g, 8 mmol) dis-
solved in THF (30 mL) with 4-allyloxybenzyl alcohol
(1.52 g, 9.3 mmol) and triphenylphosphine (2.4 g, 9.3
mmol) was added diethylazodicarboxylate (1.6 g, 9.3
mmol). The mixture was warmed to 25 ꢀC, the solvents
evaporated, and the residue puri®ed using hexane:ethyl
acetate (5:1) to give 3.85 g (75%) of 8a.
20f. To a solution of 18 (0.05 g, 0.047 mmol) in DMF
(0.5 mL) and triethylamine (6.6 mL, 0.047 mmol) was
added benzoyl chloride (0.0107 g, 0.047 mmol). After 3
h the reaction was complete and the solvents were eva-
porated and the residue puri®ed using 10% methanol/
DCM/1% formic acid to give 29 mg (50%) 20f.
20g. To a solution of 18 (0.025 g, 0.024 mmol) in DMF
(0.5 mL) and triethylamine (4 mL, 0.028 mmol) were
added dimethylaminopyridine (DMAP) (0.0033 g, 0.028
mmol) and succinic anhydride (0.0026 g, 0.026 mmol).
After 1 h the reaction was complete and the solvents
were evaporated and the residue was puri®ed using 5%
methanol/DCM/1% formic acid to give 15 mg (57%)
20g.
Fmoc-Homotyrosine 40-carboxymethylphenylmethyl ether
TSE ester (8b). A solution of 8a (2.5 g, 3.8 mmol) in
methanol:DCM (1:1) (20 mL) cooled to 78 ꢀC was
bubbled in ozone for 20 min. Dimethyl sul®de was added
and the solvents were evaporated. This crude residue was
dissolved in 50% aq THF (50 mL) and combined with 2-
methylbut-2-ene (9.4 mL, 2 M in THF), a freshly pre-
pared solution of sodium chlorite (80%) and potassium
phosphate. After 3 h the mixture was diluted with ethyl
acetate and washed with 10% sodium bisul®te, brine
and the solvents were evaporated. The residue was pur-
i®ed using hexane:ethyl acetate (3:1±1:2 with 1% acetic
acid) to give 2.36 g (92%) of 8b.
20h. To a solution of 18 (0.1 g, 0.096 mmol) in DMF (1
mL) and triethylamine (40 mL) were added N,N0-bis-t-
butoxycarbonylthiourea (0.029 g, 0.105 mmol) and
mercuric chloride (29 mg, 0.105 mmol). After 2 h the
solvents were evaporated, and the crude residue was
puri®ed using 15% methanol/DCM to give 116 mg bis-
Boc intermediate which was directly treated with 90%
tri¯uoroacetic acid (3 mL) for 3.5 h. The solvents were
evaporated and the residue puri®ed by HPLC to give 11
mg (65%) of 20h.
a. Preparation of resin 9b. (i) Fmoc-9b: the hTyr-linker
portion was loaded onto the resin in the following
manner:25 to a solution of 8b in DCM:N-methylpyrrol-
idone (NMP) (40 mL, 3:1) was added HOBT (1.2 equiv)
at 0 ꢀC in a separatory funnel followed by DIPC (1.2
equiv). After 30 min this mixture was added to base-
washed BHA resin (Novabiochem; 0.6±0.9 mmol/g) and
rinsed with more DCM (10 mL). This mixture was sha-
ken overnight and washed the following day with DCM
(3Â50 mL for 5 min) followed by MeOH (3Â50 mL for 5
min). Endcapping was carried out by shaking the resin
with acetic anhydride (4 equiv/mol of resin) overnight
followed by washing with DCM (3Â50 mL for 5 min),
DMF (3Â5 mL for 5 min), and MeOH (3Â5 mL at m
for 5 min) to give resin after drying by vacuum eva-
poration. The loading number was calculated to be 0.46
mmol/g of resin Fmoc-9b.
20i. To a solution of 18 (0.04 g, 0.038 mmol) in DMF
(0.5 mL) and l,2-dichloroethane (1 mL) were added
TEA (0.011 mL) and Boc2O (12.3 mg, 0.057 mmol) and
the mixture was stirred for 17 h. The solvents were eva-
porated and the residue puri®ed using 10% methanol/
DCM to give 31.4 mg (73%) 20i.
20j. To a solution of 18 (0.03g. 0.028 mmol) in DMF
(0.5 mL) was added a solution of 1 M methyl chloro-
formate in DCM (0.03 mL, 0.029 mmol) and the mix-
ture was stirred for 2 h. The solvents were evaporated
and the residue puri®ed using 15% methanol/DCM to
give 24 mg (90%) 20j.
(ii) Standard Fmoc deprotection the resin Fmoc-9b
from above was washed with DMF (2Â75 mL at m
for 5 min), treated with 20% piperidine in DMF
(2Â75 mL at m for 15 min) and washed with DMF
(6Â50 mL for 5 min) to give 9b ready for coupling
reactions.
20k. To a solution of 18 (0.053g, 0.05 mmol) in DMF
(0.5 mL) was added a solution of 1 M methyl iso-
cyanine in DCM (0.05 mL) and the mixture was stirred
overnight. The solvents were evaporated and the residue

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
1691
b. Preparation of pentapeptide 15b. (i) To resin 9b (10 g,
5 mmol) in NMP (40 mL) were added FmocHyp(t-
BuO)OH (9.4 g, 20 mmol), HOBT (3 g, 22 mmol), and
DIPC (3.46 mL, 22 mmol) and the mixture was shaken
overnight. The resulting resin was ®ltered. Standard
washing procedure: the resin was washed consecutively
with DMF (3Â50 mL for 5 min), DCM (3Â50 mL for 5
min), and DMF:DCM (3Â50 mL, 1:1, for 5 min).
before puri®cation by HPLC or ¯ash chromatography
on silica gel to give the ®nal cyclic hexapeptide, 21.
The following Boc-protected amino acids were used as
Thr replacements, and the total yield from 15b for (i/ii)
and (iii) were, respectively: (l)-BocSer(t-Bu)OH, 21a,
8%, 68%; (l)-Boc-ethylGlyOH, 21b, 8%, 26%; (l)-Boc-
GlyOH, 21c, 15%, 64%; (d)-BocThr(t-Bu)OH, 21d, 9%,
28%; (l)-Boc-a,a-dimethylGlyOH, 21e, 13%, 37%; (l)-
BocTyr(t-Bu)OH, 21f, 13%, 42%; (l)-BocGlu(t-Bu)OH,
21g, 21%, 32%; (l)-BocOrn(Z)OH, 21h, 8%, 27%; (l)-
BocDap(Z)OH, 21i,13%, 38%; (l)-BocArg(di-Z)OH,
21j, 11%, 32%.
(ii) Resin from (i) above was deprotected as in a(ii) for
Fmoc-9b followed by combining FmocThr(Ot-Bu)OH
(8 g, 20 mmol) in NMP (50 mL), HOBT (3 g, 22 mmol),
DIPC (3.46 mL, 22 mmol) and the mixture was shaken
overnight. The resulting resin was washed as in b(i).
Section C. Preparation of hTyr analogues (28a±h)
(iii) One half of the resin from above was deprotectcd as
in a(ii) above followed by combining FmocOrn(Aloc)
OH (2.2 g, 2.5 mmol) in NMP (50 mL), HOBT (0.76 g,
5.5 mmol), DIPC (0.87 mL, 5.5 mmol) and the mixture
was shaken overnight. This resin was washed as in b(i)
and dried to give the peptide resin 14b.
a. Preparation of homotyrosine (hTyr) replacement acids
(26a±h). Homotyrosine (hTyr) replacement acids 26 for
couplings were either commercially available (a,d,e,f) or
prepared as follows:
26b. (i) To a solution of BocTyrOMe (23b) (2.5 g, 8.47
mmol) and PhN(Tf)2 (3.18 g, 8.47 mmol) in DCM (12
mL) at 0 ꢀC was added TEA (1.3 mL) and the mixture
was warmed to 25 ꢀC overnight. The mixture was washed
with 1 M NaHCO3, and the solvents were evaporated.
The crude residue was puri®ed using ethyl acetate:hexane
(1:4) to give 3.2 g (88%) oil BocTyr (OTf)OMe. A solu-
tion of this tri¯ate (2 g, 4.68 mmol), potassium acetate
(2.3 g, 23.4 mmol), Pd(OAc)2 (0.105 g, 0.47 mmol),
DPPF (0.52 g, 0.94 mmol) in DMF (10 mL) was heated
at 60 ꢀC for 2 days. The mixture was combined with
ethyl acetate and satd NaHCO3 and ®ltered. The aqu-
eous layer was separated, acidi®ed with 10% citric acid,
extracted with ethyl acetate and the organic layer was
separated and the solvent was evaporated to give 0.62 g
(41%) of the crude solid (24b).
(iv) Resin 14b (10 g, 4.6 mmol) was deprotected as in
a(ii) and reacted with 2,4,5-trichlorophenyl 400-pentyl-
oxyterphenoate (5 g, 9.2 mmol) and DIPEA (1.42 g, 5.5
mmol) in NMP (50 mL). This mixture was shaken at
60 ꢀC overnight. The next day it was washed as in b(i) to
give acylated resin. The acylated resin was treated with
tri-n-butyltin hydride (4 g, 13.8 mmol), HOAc (1.5 mL),
DCM (40 mL), and Pd(Ph3P)4 (0.22 g, 0.18 mmol).
After shaking overnight this resin was washed with
DCM (2Â50 mL for 5 min), 10% TEA in DCM (2Â50
mL for 5 min), DCM (2Â50 mL for 5 min), MeOH
(2Â50 mL for 5 min) and DCM (2Â50 mL for 5 min) to
give dried resin 15b.
c. Preparation of residue B analogues (21a±j). (i) Resin
15b was treated with 13 (4.4 equiv), HOBT (4 equiv),
DIPC (4 equiv) in NMP. After shaking overnight resin
16b was washed as in b(i), deprotected as in a(ii), and
treated with a Boc-protected threonine (Thr) replace-
ment (see Table 3), HOBT (4 equiv), and DIPC (4
equiv) in NMP. After shaking overnight this resin was
washed as in b(i) to give 17b.
(ii) To a solution of 24b (0.62 g, 1.91 mmol) in THF (10
mL) at 0 ꢀC was added 10 M borane:dimethyl sul®de
complex (0.9 mL, 8.9 mmol) and the mixture was
warmed to 25 ꢀC for 2 h. The solvents were evaporated
and the residue partitioned between ethyl acetate and
satd NaHCO3. The organic layer was separated and
evaporated. The crude residue was puri®ed using ethyl
acetate:hexane (2:3) to give 0.45 g (76%) BocTyr(CH2
OH)OMe (25b).
(ii) The peptide was cleaved from resin 17b by shaking
with 90% TFA for 2±3 h. The resin was ®ltered, rinsed
with 90% TFA, and the TFA was evaporated. The
residue was triturated with ether, and the solids were
®ltered and dried. The crude residue was dissolved in
DMF (10 mL) and combined with HOBT (2.5 equiv),
DIPEA (5 equiv), and PyBOP (2.5 equiv) and stirred
overnight. The next day the solvents were evaporated, the
residue was dissolved in MeOH and precipitated by add-
ing water. The crude solids were ®ltered and dried to give
the N-protected cyclic hexapeptide which was puri®ed
using 5±15% methanol in DCM.
(iii) To a solution of BocTyr(CH2OH)OMe (0.43 g, 1.39
mmol) in THF (2.5 mL) at 0 ꢀC was added lithium
hydroxide (0.07 g, 1.67 mmol) in water (1.5 mL) and
stirred for 2 h. The mixture was diluted with ethyl ace-
tate, the aqueous layer was separated, acidi®ed with
10% citric acid, and extracted with EtOAc. Evaporation
of the solvent gave 0.4 g (97%) of 26b.
26c. (i) To a solution of hTyr (2.3 g, 8.3 mmol) in
MeOH (20 mL) was added thionyl chloride (1.18 g, 10
mmol) and the mixture was re¯uxed for 6 h. The sol-
vents were evaporated and the crude residue was pur-
i®ed using 15% MeOH/DCM to give 2 g (98%) of
methyl ester. To a solution of ester (2 g, 8.1 mmol)
in dioxane (80 mL) was added 1 M NaHCO3 (26 mL)
(iii) This hexapeptide was shaken with 10% Pd/C for 2±
3 h in ethanol with a hydrogen balloon, and the mixture
was ®ltered, rinsed with more MeOH, and the washings
were collected and evaporated. The residue was tritur-
ated with ether, the crude solids were ®ltered and dried

1692
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
followed by Boc2O (2.43 mL, 10.5 mmol) and the mix-
ture was stirred overnight. The solvents were evapo-
rated and the residue was partitioned between ethyl
acetate and water, the organic layer was separated and
washed with 10% citric acid, brine, and the solvent was
evaporated. The residue was puri®ed using 5% MeOH/
DCM to give 1.7 g (67%) of Boc-hTyrOMe (23c).
bubbled in gaseous CO and the mixture was heated to
60 ꢀC for 8 h. The solvents were evaporated and the
crude residue was puri®ed using 50% ethyl acetate/hex-
anes to give 0.8 g (87%) acid 24g.
(iii) To a solution of acid from (ii) (0.8 g, 1.9 mmol) in
THF (12 mL) was added 10 M BH3S(CH3)2 (0.97 mL,
9.7 mmol) and the mixture was stirred for 4 h. The
mixture was cooled to 0 ꢀC and quenched with MeOH
(10 mL). The solvents were evaporated and the crude
residue was puri®ed using 35% ethyl acetate/hexanes to
give 0.33 g (43%) of the alcohol 25g.
(ii) To a solution of 23c (1.7 g, 5.5 mmol) was added
PhN(Tf)2 (2.16 g, 5.8 mmol) and TEA (0.85 mL, 6.1
mmol) and the mixture was stirred overnight. The sol-
vent was evaporated and partitioned between 1 M
NaHCO3 and ether, and the organic layer was separated
and evaporated. The crude residue was puri®ed using
15% MeOH/DCM to give 2 g (82%) of the tri¯ate. A
solution of tri¯ate (2 g, 4.53 mmol), potassium acetate
(2.2 g, 25 mmol), Pd(OAc)2 (0.3 g, 1.36 mmol), DPPF (1
g, 1.8 mmol) in DMF (10 mL) was heated at 60 ꢀC for
16 h. The mixture was combined with ethyl acetate and
satd NaHCO3 and ®ltered. The aqueous layer was
separated, acidi®ed with 10% citric acid, extracted with
ethyl acetate and the organic layer was separated and
the solvent was evaporated to give 1.25 g (82%) of the
aryl acid 24c.
(iv) A solution of alcohol (0.33 g, 0.83 mmol) in MeOH
(5 mL) and 10% Pd/C (0.05 g) was hydrogenated with a
hydrogen balloon for 1.5 h. The mixture was ®ltered
and the crude residue was puri®ed using 20% MeOH/
DCM to give 0.175 g (73%) of the methyl analogue 26g.
26h. To a solution of 24c (0.35 g, 1 mmol) in THF (20
mL) was added DPPA (0.246 mL, 1.1 mmol) followed
by TEA (0.16 mL, 1.1 mmol) and the mixture was
heated at 60 ꢀC for 2 h. The solvents were evaporated,
the residue was dissolved in toluene (5 mL) and benzyl
alcohol (0.139 mL, 1.3 mmol) was added and the mix-
ture was heated to 110 ꢀC for 4 h. The solvents were
evaporated and the crude residue was puri®ed using
25% ethyl acetate/hexanes to give 0.28 g (61%) benzyl
carbamate 25h. To a solution of carbamate (0.28 g, 0.63
mmol) in THF (2 mL) at 0 ꢀC was added LiOH (0.032 g,
0.76 mmol) in water (1 mL) and the mixture was stirred
for 3 h. The solvents were evaporated and after the
usual work up the crude residue was puri®ed using 10%
MeOH/DCM to give 0.27 g (99%) acid 26h.
(iii) To a solution of acid (0.7 g, 2 mmol) in THF (20
mL) was added 10 M borane:dimethyl sul®de complex
(1 mL, 10 mmol) and the mixture was stirred for 12 h.
The solvents were evaporated and the residue parti-
tioned between ethyl acetate and satd NaHCO3, the
organic layer was separated and evaporated. The crude
residue was puri®ed using 35% ethyl acetate/hexane to
give 0.51 g (76%) Boc-hTyr(CH2OH)OMe (25c).
(iii) To a solution of Boc-hTyr(CH2OH)OMe (0.51 g,
1.57 mmol) in THF (2.5 mL) at 0 ꢀC was added lithium
hydroxide (0.08 g, 1.9 mmol) in water (1.5 mL). The
mixture was warmed to 25 ꢀC over 1.5 h and the sol-
vents were evaporated. The residue was partitioned
between water and ethyl acetate, and the aqueous layer
was separated, acidi®ed with 10% citric acid, extracted
with ethyl acetate, and the organic layer was separated
and evaporated to give 0.47 g (96%) Boc-hTyr(CH2
OH)OH (26c).
b. Preparation of pentapeptide (22). (i) To a solution of
ProOt-Bu (5 g, 29.2 mmol) and FmocThr(t-BuO)OH
(12.8 g, 32.2 mmol) in DCM (150 mL) at 0 ꢀC were
added EDAC (8.6 g, 45 mmol) and HOBT (5.8 g, 43
mmol). The mixture was warmed to 25 ꢀC overnight.
The solvent was evaporated and the residue was com-
bined with ethyl acetate and 10% citric acid. The
organic layer was washed with satd Na2CO3, and dried
with Na2SO4. The solvent was evaporated to give crude
solids which were used directly for the next step.
26g. (i) Boc-hTyr (2.1 g, 7 mmol) was dissolved in
MeOH (40 mL) and water (4 mL) and 20% CsCO3 was
added to pH 7 and the solvents were evaporated. The
residue was dissolved in DMF (20 mL), and benzyl
bromide (0.88 mL) was added and the mixture was stir-
red for 16 h. After the standard work up, the crude
residue was puri®ed using 5% MeOH/DCM to give 0.95
g (35%) (23g).
(ii) This dipeptide was dissolved in DMF (50 mL) and
diethylamine (15 mL) for 4 h. The solvents were
removed and the crude residue was combined with
FmocOrn(Z)OH (15.7 g, 32 mmol), EDAC (6.8 g, 36
mmol), and HOBT (4.8 g, 36 mmol) and stirred over-
night. The solvent was removed and the crude residue
was combined with ethyl acetate and 10% citric acid.
The organic layer was washed with satd NaHCO3, and
dried with Na2SO4. The solvent was evaporated and the
crude residue was puri®ed using 2.5% MeOH in DCM
to give 11 g (47% from ProOt-Bu) of FmocOrn(Z)-
Thr(t-OBu)-ProOt-Bu.
(ii) To a solution of the benzyl ester (0.85 g, 2.2 mmol)
in DCM (10 mL) at 0 ꢀC were added PbNTf2 (0.87 g,
2.37 mmol) and TEA (0.34 mL), and the mixture was
stirred for 16 h. The solvents were evaporated and the
crude residue was puri®ed using 15% ethyl acetate/hex-
anes to give 1.13 g (99%) tri¯ate. To a suspension of
tri¯ate from (ii) (1.13 g, 2.22 mmol), potassium acetate
(1.09 g, 11.1 mmol), palladium acetate (0.05 g, 0.22
mmol), DPPF (0.25 g, 0.44 mmol) in DMF (10 mL) was
(iii) Tripeptide from (ii) (2.67 g, 3.3 mmol) was treated
with diethylamine (4 mL) in DMF (20 mL) for 4 h. The
solvents were evaporated and the crude residue was
dissolved in DMF (40 mL) and treated with DIPEA

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
1693
(1.74 mL) and 2,4,5-trichlorophenyl 4-(N-piperazinyl-
N0-(4-n-octyloxyphenyl)benzoate (PIP8) (for b,c,g,h) or
2,4,5-trichlorophenyl 4-((4-n-pentyloxyphenyl)-4-phe-
nyl)benzoate (Ter) (for a,d,e,f) (2.16 g, 3.7 mmol) and
heated to 50 ꢀC for 2 days. The solvents were evapo-
rated, and the crude residue was puri®ed using ethyl
acetate:hexane (3:1) to give 1.39 g (43% from (ii)) of
PIP8Orn(Z) Thr(t-OBu)ProOt-Bu.
for 1 h. The solvents were evaporated and the crude
residue was puri®ed using 7.5±10% MeOH/DCM with
1% NH4OH. The material was dissolved in MeOH and
excess acetic acid, and the solvents were evaporated to
give the acetate salt 28.
28a. For example, the yields of coupling of 22a and 26a,
TFA treatment and cyclization of 27a, and TPP reduc-
tion of azide as described above giving 28a were 98, 55
and 63%, respectively.
(iv) A solution of tripeptide from (iii) above (2.6 g, 2.7
mmol) in ethanol (50 mL) and acetic acid (0.35 mL) was
treated with Pd(OH)2 (0.44 g) with a hydrogen balloon.
After 5 h the catalyst was ®ltered and the solvent was
evaporated. The residue was partitioned between ethyl
acetate and satd NaHCO3, the organic layer was sepa-
rated, washed with brine and dried with Na2SO4. The
solvent was evaporated to give 2 g crude amine. This
crude amine was combined with FmocAzpOH (13) (1.11
g, 2.9 mmol), CMC (1.7 g, 4 mmol), and HOBT (0.54 g, 4
mmol) in DCM (30 mL). After the standard work up the
residue was puri®ed using 3% MeOH/DCM to give 1.87
g (60%) of the tetrapeptide.
28b. The same reaction for 22b and 26b, 27b, resulting
in 28b, gave 57, 30, 94%.
28c. The same reaction for 22b and 26c, 27c, resulting in
28c, gave 73, 19, 77%.
28d. The same reaction for 22a and 26d, 27d, resulting
in 28d, gave 99, 71, 60%.
28e. The same reaction for 22a and 26e, 27e, resulting in
28e, gave 99, 37, 67%.
(v) To a solution of the tetrapeptide (1.86 g, 1.56 mmol)
in acetonitrile (12 mL) at 0 ꢀC was added diethylamine
(3 mL) and the mixture was warmed to 25 ꢀC for 2 h.
The solvents were evaporated, and the residue was
combined with FmocThr(t-BuO)OH (0.7 g, 1.8 mmol),
EDAC (0.313 g, 1.6 mmol), HOBT (0.25 g, 1.8 mmol) in
NMP (25 mL) at 0 ꢀC before adding NMM (0.2 mL).
The mixture was stirred at 25 ꢀC overnight, and the sol-
vents were evaporated. The crude residue was puri®ed
using 8% MeOH/DCM to give 1.8 g (85%) of the pen-
tapeptide 22.
28f. The same reaction for 22a and 26f, 27f, resulting in
28f, gave 79, 47, 67%.
28g. The same reaction for 22b and 26g, 27g, resulting
in 28g, gave 97, 26, 94%.
28h. The same reaction for 22b and 26h, 27h, resulting
in 28h, gave 51, 33, 67%.
Section D. Side-chain analogues
a. Preparation of cyclic hexapeptides via the `3+3' ap-
proach. Fmoc-hTyr-Thr-AzpOMe (29). (i) BocThrOH
(11.32 g, 52 mmol), 12 (10.67 g, 52 mmol), and HOBT
(8.72 g, 64 mmol) were dissolved in NMP (70 mL) and
THF (170 mL) and cooled to 0 ꢀC. NMM (6.95 mL)
was added followed by addition of EDAC (10.84 g, 57
mmol), and the reaction mixture was allowed to warm
to 25 ꢀC overnight. After the standard work up, the
residue was redissolved in ethyl acetate, precipitated in a
mixture of 10% MTBE in heptane, stirred for 30 min at
0 ꢀC, ®ltered and dried in vacuo to give 15.47 g (75%)
BocThrAzpOMe.
c. General procedure for preparation of the hTyr analo-
gues (28a±h). (i) To a solution of the pentapeptide
22a,b (1 g, 0.74 mmol) in acetonitrile (9 mL) at 0 ꢀC was
added diethylamine (2 mL) and the mixture was
warmed to 25 ꢀC for 2 h. The solvents were evaporated,
and the residue was puri®ed using 8% MeOH/DCM to
give 0.77 g (92%) of amine. To a 0 ꢀC solution of this
amine (0.2 g, 0.17 mmol) in DCM (2 mL) were added
CMC (3 equiv), HOBT (3 equiv) and the hTyr replace-
ment (1.25±2 equiv), and this mixture was warmed to
25 ꢀC overnight. The solvents were evaporated and after
the standard work up the crude residue was puri®ed
using 5% MeOH/DCM to give the acyclic hexapeptide
27a,b.
(ii) Treatment of BocThrAzpOMe (12.35 g, 33 mmol)
with HCl/ethyl acetate as for 12 gave 9.93 g (100%)
ThrAzpOMe(HCl).
(ii) Hexapeptide 27a,b was treated at 0 ꢀC with 90% aq
tri¯uoroacetic acid and then warmed to 25 ꢀC for 3 h.
The mixture was precipitated with diethyl ether and the
solids ®ltered to give the crude amino acid. To a solu-
tion of this material dissolved in DMF at 0 ꢀC was
added DPPA (1.1 equiv). The mixture was warmed to
25 ꢀC overnight. The solvents were evaporated and after
the standard work up, the crude residue was puri®ed
using 5±10% MeOH/DCM to give the cyclized product.
(iii) ThrAzpOMe(HCl) (9.9 g, 33 mmol), Fmoc-
hTyrOH (13.52 g, 33 mmol), and HOBT (5.55 g, 41
mmol) were dissolved in NMP (150 mL) and THF (150
mL) and cooled to 0 ꢀC. NMM (5.43 mL) was added
followed by addition of EDAC (6.67 g, 35 mmol), and
the reaction mixture was kept at 0 ꢀC overnight. The
mixture was warmed to 25 ꢀC, concentrated, and after
the standard work up, the organic layer was washed
with 10% aqueous sodium bicarbonate, brine, and the
solvent was evaporated. The residue was dissolved in
ethyl acetate, precipitated in a mixture of 10% MTBE
in heptane, stirred for 30 min at 0 ꢀC, ®ltered and dried
(iii) To a solution of the cyclic hexapeptide (0.055 g,
0.048 mmol) in THF:water (4:1) was added triphenyl-
phosphine (4 equiv) and the mixture was heated at 80 ꢀC

1694
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
in vacuo to give 21.04 g (97%) Fmoc-hTyrThr-Pro
(N3)OMe (29).
(iii) Hexapeptide 31 (9 g, 6.8 mmol) was dissolved in
90% aq ethanol (150 mL) and cooled to 0 ꢀC. To this
solution was added 1 N sodium hydroxide (35 mL)
keeping the pH at 12 at 0±25 ꢀC until the hydrolysis was
complete. The pH was adjusted to 5, the ethanol eva-
porated, and the aqueous layer was extracted with ethyl
acetate to remove impurities. The aqueous layer was
evaporated and the residue was dissolved in methanol,
salts ®ltered, and the solvent was evaporated and crude
residue dried in vacuo. The crude residue was dissolved
in dimethylformamide (1.3 L) and cooled to 20 ꢀC.
Diphenylphosphoryl azide (2.2 mL, 10.2 mmol) was
added with stirring followed by sodium bicarbonate (4.1
g). The reaction mixture was kept at 25 ꢀC until the
coupling was complete. The solvent was evaporated and
the residue was dissolved in 10% methanol/ethyl ace-
tate. Precipitated solids were ®ltered and the ®ltrate was
washed with 10% aq sodium bicarbonate and brine.
The solvents were evaporated and the crude material
was puri®ed using 8% methanol/DCM to give 5 g
(69%) of the cyclic hexapeptide 32.
BocOrn(Fmoc)ThrHypOH (30). (i) CbzThrOH (22 g, 87
mmol), HypOBn(HCl) (22.39 g, 87 mmol), and HOBT
(14.81 g, 109 mmol) were dissolved in NMP (240 mL)
and THF (240 mL) and cooled to 0 ꢀC. NMM (14.4 mL)
was added followed by addition to EDAC (17.7 g, 93
mmol), and the reaction mixture was allowed to warm to
25 ꢀC overnight. The mixture was concentrated, and
after the standard work up, the residue was dissolved in
ethyl acetate, precipitated in a mixture of 10% MTBE in
heptane, stirred for 30 min at 0 ꢀC, ®ltered and dried in
vacuo to give 37.5 g (95%) CbzThrHypOBn.
(ii) To a solution of CbzThrHypOBn (22 g, 48 mmol) and
p-toluenesulfonic acid (10.09 g, 53 mmol) in isopropanol
(200 mL) was added 5% Pd/C and the mixture was
hydrogenated via hydrogen balloon pressure for 16 h. The
catalyst was ®ltered and washed with methanol and the
solvents were evaporated to give a residue ThrHyp
OH(TsOH) which was used directly for the next step.
(iii) To a 0 ꢀC solution of BocOrn(Fmoc)OH (25 g, 55
mmol) and N-hydroxysuccinimide (HOSu) in THF (300
mL) was added a solution of DCC (12.46 g, 60 mmol) in
THF and the reaction was stored at 0 ꢀC overnight. The
solids were ®ltered, the ®ltrate concentrated, and com-
bined with heptane to precipitate a white solid which
was ®ltered and rinsed with more heptane. The crude
residue was dissolved in water (110 mL) with sodium
bicarbonate (10 g) and cooled to 0 ꢀC. A solution of
BocOrn(Fmoc)OSu was added with stirring and the
mixture was allowed to warm to 25 ꢀC for 20 h. The
THF was evaporated and the pH brought to 7. The
mixture was diluted with water and extracted with ethyl
acetate. The organic layer was discarded and the pH of
the aqueous layer was adjusted to 4 with 20% citric acid
and extracted with ethyl acetate. The solvents were
evaporated and the crude residue was puri®ed using
ethyl acetate:methanol:acetic acid (100:5:1±100:10:3).
The product was then precipitated from hexanes:diethyl
ether (4:1) to give 28.57 g (88%) of the tripeptide
BocOrn(Fmoc)ThrHypOH (30).
b. Preparation of side-chain analogues (34a±k). To a
solution of 32 (4.1 g, 4.7 mmol) in chloroform (4 mL) at
0 ꢀC was added 90% TFA (14 mL) and the mixture
allowed to warm to 25 ꢀC. After stirring for 2 h ether
was added to precipitate the solids which were ®ltered,
rinsed with ether and vacuum dried to give 4 g (99%) of
the tri¯uoroacetic acid salt of 33a.
34a. To a solution of 33a (50 mg, 67 mmol) in THF (4
mL) and DIPEA (24 mL, 0.17 mmol) was added acetic
anhydride (8 mg, 0.08 mmol). After 2 days the solvents
were evaporated and the crude residue was puri®ed using
20% methanol in DCM with 5% ammonium hydroxide
to give 33 mg (62%) of the N-acetyl intermediate. This
material was treated with triphenylphosphine (40 mg,
0.15 mmol) in 20% aq THF (6 mL) at re¯ux for 2 h. The
solvents were evaporated and the crude residue was pur-
i®ed by using 30% methanol in DCM with 5% ammo-
nium hydroxide to give 28.7 mg (99%) of 34a.
34b. To a solution of 33a (72 mg, 0.098 mmol) in DMF (4
mL) and DIPEA (31 mg, 0.24 mmol) was added 2,4,5-
trichlorophenyl pentadecanoate (83 mg, 0.19 mmol).
After 2 days the solvents were evaporated and the crude
residue was precipitated with 20% methanol in DCM
and ether to give 54 mg azide which was hydrogenated
with 5% Pd/C (20 mg) in ethanol (5 mL) equipped with a
hydrogen balloon. The catalyst was ®ltered, the solvents
were evaporated and the crude residue was puri®ed by
using 20% methanol in DCM with 5% ammonium
hydroxide to give 48 mg (53%) of 34b.
Preparation of cyclic hexapeptide (32). (i) A solution of
FmocHtyThrAzpOMe (29) (6.73 g, 10 mmol) in aceto-
nitrile (150 mL) was cooled to 0 ꢀC and diethylamine (35
mL) was added and the mixture was stirred for 1 h. The
solvent was evaporated and the crude residue (4.6 g)
was used directly without further puri®cation.
(ii) To a solution of BocOrn(Fmoc)ThrHypOH (30)
(5.22 g, 8 mmol) and crude HtyThrAzpOMe from
above in DCM (150 mL) was added a solution of
EEDQ (2.73 g, 11 mmol) and the mixture stirred for 40
h. The reaction mixture was ®ltered through Celite and
the solvents were evaporated. After the standard work up,
the crude residue was puri®ed using 4±10% methanol in
ethyl acetate to give 9.2 g (70%) of the linear hexapeptide
BocOrn(Fmoc)ThrHypHtyThrAzpOMe (31).
34c. To a solution of 33a (75 mg, 0.1 mmol) in THF (4
mL) and DIPEA (34 mL, 0.25 mmol) was added benzoic
anhydride (27 mg, 0.12 mmol). After 1 h the solvents
were evaporated and the crude residue was puri®ed
using 20% methanol in DCM with 5% ammonium
hydroxide to give 70 mg (82%) of the N-benzoate. This
material was treated with triphenylphosphine (86 mg,

L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
1695
0.33 mmol) in 20% aq THF (6 mL) at re¯ux for 4 h.
The solvents were evaporated and the crude residue was
puri®ed by using 15% methanol in DCM with 5%
ammonium hydroxide to give 28.7 mg (67%) of 34c.
35e. Compound 35e was prepared in a similar manner
to 35c by replacing the residue B BocDap(Cbz)OH with
BocOrn(Cbz)OH. Reduction of the azido and Cbz
groups was carried out with hydrogen balloon, 10% Pd
on carbon in THF.
34d,f±k. Side-chains used to acylate 33a to give 34d±k
were all reacted as the corresponding 2,4,5-tri-
chlorophenyl esters (1.5 equiv) with DIPEA (2.5 equiv)
in DMF at 50 ꢀC until complete. The solvents were eva-
porated and the crude residues were puri®ed using 10±
20% methanol/DCM with 5% ammonium hydroxide to
give the pure azidoproline precursors 34d±k, all of which
except 34e (see below) were reduced by heating with TPP
(4 equiv) in 20% aq THF until complete. The solvents
were evaporated and the crude residues were puri®ed
with similar solvents to give the ®nal 34d,f±k.
35f. Compound 35f was prepared in a similar manner to
35b. Deprotection of the Cbz group was carried out
using hydrogen balloon, 20% Pd(OH)2 on carbon in
THF in the presence of AcOH.
35g. Compound 35g was prepared in a similar manner
to 34i by replacing the residue B BocThrOH with
BocOrn(Cbz)OH and replacing the residue D HypOBn
(HCl) with ProOBn(HCl). Sequential reduction of the
azide with Ph3P in THF±H2O and deprotection of the
Cbz group with hydrogen balloon, 20% Pd(OH)2 on
carbon in EtOH in the presence of AcOH gave 35g.
34e. In the case of the azido-precursor to 34e, the pro-
duct from above was debenzylated and reduced by
treating 33 (R=COPh-p-(N-piperazine)) (70 mg, 0.065
mmol) with 10% Pd/C (70 mg) in methanol (5 mL)
equipped with a hydrogen balloon and stirred over-
night. The catalyst was ®ltered, the solvents evaporated,
and the residue puri®ed using 10% methanol in DCM
with 2% ammonium hydroxide to give 35 mg (51%) of
34e.
35h. Compound 35h was prepared in a similar manner
to 35b. Deprotection of the Cbz group was carried out
using hydrogen balloon, 20% Pd(OH)2 on carbon in
EtOH in the presence of AcOH.
Acknowledgements
Side-chains for 34f±k were prepared as in ref 20 via
alkylation of the phenolic group. In a similar manner,
the biphenyl side-chain for 34d was prepared by octyla-
tion of methyl 40-hydroxy-4-biphenyl-carboxylate as in
this reference.
We acknowledge the support of biologists Drs. R.
Goldman, D. Frost, and J. Fostel toward this work. We
are especially appreciative of the D-4MP chemists Drs.
Q. Li, W.-B. Wang, A. Claiborne, L. Hasvold, C. Berst
and B. Steiner who helped to carry out these initial
results. We greatly appreciate the e€orts of Dr. Jaroslav
Kyncl to obtain large quantities of starting materials
from several sources. We are grateful to Dr. Paul West
(D-418) for his acquisition of the high resolution mass
spectral data.
Section E. Preparation of combination analogues (35a±h)
35a. Compound 35a was prepared in a similar manner to
18 by replacing the terphenyl side-chain with 4-(N-(4-(n-
pentyloxy)phenyl)piperazinyl)benzoyl group and the
residue B threonine with BocDap(Cbz)OH. For this
compound, the cyclization was accomplished using
DPPA in the presence of NaHCO3 as described in 32 (iii).
Sequential reduction of the azide with Ph3P in THF±
H2O and deprotection of the Cbz group with hydrogen
balloon, 10% Pd on carbon in THF:EtOH (1:1) and
AcOH (a few drops) gave 35a.
References and Notes
1. Boutati, E. I.; Anaissie, E. J. Blood 1997, 90, 999.
2. Gallis, H. A.; Drew, R. H. Rev. Infect. Dis. 1990, 12, 308.
3. Rex, J. H.; Rinaldi, M. G.; Pfaller, M. A. Antimicrob.
Agents Chemother. 1995, 39, 1. Alexander, B. D.; Perfect, J. R.
Drugs 1997, 54, 657.
4. Quigley, D. R.; Selitrenniko€, C. P. Exp. Mycol. 1984, 8, 202.
5. (a) Frost, D. J.; Brandt, K.; Capobianco, J. O.; Goldman,
R. C. Microbiology 1994, 140, 2239. (b) Capobianco, J. O.;
Zakula, D.; Frost, D. J.; Goldman, R. C.; Li, L.; Klein, L. L.;
Lartey, P. A. Antimicrob. Agents Chemother. 1998, 42, 389.
6. Gordee, R. S. In New Approaches for Antifungal Drugs;
Fernandes, P. B., Ed.; Birkhauser: Boston, 1992.
35b. Compound 35b was prepared from the amino pre-
cursor of 35a. After reduction of the azide with Ph3P in
THF±H2O, the free amine was guanidinylated as for
20h. Deprotection of the Cbz group was carried out as
described above to give 35b.
7. (a) Keller-Juslen, C.; Kuhn, M.; Loosli, H. R.; Petcher, T.
J.; Weber, H. P.; Von Wartburg, A. Tetrahedron Lett. 1976,
46, 4147. (b) Benz, F.; Knuesel, F.; Nuesch, J.; Treichler, H.;
Voser, W.; Nyfeler, R.; Keller-Schierlein, W. Helv. Chim. Acta
1974, 57, 2459. (c) Reviews: Turner, W. W.; Rodriguez, M. J.
Curr. Pharmaceut. Des. 1996, 2, 209. Balkovec, J. M. Expert
Opin. Invest. Drugs 1994, 3, 65. Debono, M. Expert Opin.
Invest. Drugs 1994, 3, 65.
35c. Compound 35c was prepared in a similar manner
to 18 by replacing the terphenyl side-chain with octyl-
oxyphenylpiperazinylbenzoyl group and the residue B
threonine with BocDap(Cbz)OH. Reduction of the
azido and Cbz groups was carried out using hydrogen
balloon, 10% Pd on carbon in THF:EtOH (1:1) and
AcOH.
8. (a) Traxler, P.; Gruner, J.; Auden, J. A. J. Antibiot. 1977,
30, 289. (b) Analogue program from this lab: Yeung, C. M.;
Klein, L. L.; Lartey, P. A. Bioorg. Med. Chem. Lett. 1996, 6,
819.
35d. Compound 35d was prepared in a similar manner
to 35b by replacing the pentyloxy side-chain with octyl-
oxyphenylpiperazinylbenzoyl group.

1696
L. L. Klein et al. / Bioorg. Med. Chem. 8 (2000) 1677±1696
9. Debono, M.; Turner, W. W.; LaGrandeur, L.; Burkhardt,
F. J.; Nissen, J. S.; Nichols, K. K.; Rodriguez, M. J.; Zweifel,
M. J.; Zeckner, D. J.; Gordee, R. S.; Tang, J.; Parr, T. R. J.
Med. Chem. 1995, 38, 3271.
10. Del Poeta, M.; Schell, W. A.; Perfect, J. R. Antimicrob.
Agents Chemother. 1997, 41, 1835. Further optimization led
to: Journet, M.; Cai, D.; DiMichele, L. M.; Hughes, D. L.;
Larsen, R. D.; Verhoeven, T. R.; Reider, P. J. J. Org. Chem.
1999, 64, 2411. Bou€ard, F. A.; Dropinski, J. F.; Balkovec, J.
M.; Black, R. M.; Hammond, M. L.; Nollstadt, K. H.; Drei-
korn, S. In Abstracts of the 36th Interscience Conference on
Antimicrobial Agents and Chemotherapy; American Society for
Microbiology, Washington, DC, 1996, F-27.
15. Meulbroek, J. A.; Oleksijew, A.; Tovcimak, A.; Alder, J.
D.; Tanaka, S. K.; DeGoey, D. A.; Klein, L. L.; Li, L.; Lartey,
P. A. In Abstracts of the 37th Interscience Conference on Anti-
microbial Agents and Chemotherapy; American Society for
Microbiology, Washington, DC, 1997, F-82.
16. Webb, T. R.; Eigenbrot, C. J. Org. Chem. 1991, 56, 3009.
17. Klein, L. L.; Curty, C. B. Personal communication.
18. US Patent 506790, M. J. Rodriguez.
19. Bou€ard, F. A.; Zambias, R. A.; Dropinski, J. F.; Balk-
ovec, J. M.; Hammond, M. L.; Abruzzo, G. K.; Bartizal, K.
F.; Marrinan, J. A.; Kurtz, M. B.; McFadden, D. C.; Noll-
stadt, K. H.; Powles, M. A.; Schmatz, D. M. J. Med. Chem.
1994, 37, 222.
11. Zambias, R. A.; Hammond, M. L.; Heck, J. V.; Bartizal,
K.; Trainor, C.; Abruzzo, G.; Schmatz, D. M.; Nollstadt, K.
M. J. Med. Chem. 1992, 35, 2843.
12. (a) Li, L.; Thomas, S.; Grampovnik, D. J.; Klein, L. L.;
DeGoey, D. A.; Chen, H.-J.; Yeung, C. M.; Leone, C.; Curty,
C.; Frost, D.; Niius, A.; Meulbroek, J.; Lartey, P. A.; Plattner,
J. J. In Abstracts of the 37th Interscience Conference on Anti-
microbial Agents and Chemotherapy; American Society for
Microbiology, Washington, DC, 1997, F-238, p 186. (b) Klein,
L. L.; Li, L. Curr. Pharmaceut. Des. 1999, 5, 57.
13. Melillo, D. G.; Larsen, R. D.; Mathre, D. J.; Shukis, W. F.;
Wood, A. W.; Colleluari, J. R. J. Org. Chem. 1987, 52, 5143.
14. Shadomy, S.; Pfaller, M. A. Laboratory studies with anti-
fungal agents: susceptibility tests and quantitation in body
¯uids. In Manual of Clinical Microbiology, 5th edn.; Balows,
A.; Hausler, W. J.; Herrman, K. L.; Isenberg, H. D.; Shad-
omy, H. J., Eds.; American Society for Microbiology:
Washington, DC, 1991.
20. Patent: WO 96/11210 9/29/95; Fujisawa Pharmaceutical
Co., Ltd.
21. L. Li, W.-b. Wang and L. Hasvold prepared these proline
analogues.
22. These analogues were prepared by: Q. Li, W.-b. Wang, L.
Hasvold, A. Claiborne and E. Steiner.
23. Klein, L. L.; Garren, K.; Schmitt, E. D-4P3; Personal
communication. All Table 6 compounds showed solubilities
>10 mg/mL in bu€ered solutions. See reference for water
soluble phosphate prodrug analogue of hTyr: Rodriguez, M.
J.; Vasudevan, V.; Jamison, J. A.; Borromeo, P. S.; Turner, W.
W.; Bioorg. Med. Chem. Lett. 1999, 9, 1863.
24. Miller, L. C.; Tainter, M. L. Proc. Soc. Exp. Biol. Med.
1944, 57, 261.
25. Grant, G. A. Principles and practice of solid-phase pep-
tide synthesis. In Synthetic Peptides: A User's Guide; Grant,
G. A., Ed.; W. H. Freeman and Co.: New York, 1992; pp
104±119.