Synthesis and analysis of polyethylene glycol linked P-glycoprotein-specific homodimers based on (-)-stipiamide

Article abstract of DOI:10.1016/S0040-4039(01)00619-0

A series of five homodimeric polyethylene glycol (PEG) linked homodimers based on the multidrug resistance reversal agent (-)-stipiamide were made and tested for their ability to interact with P-glycoprotein, the protein responsible for multidrug resistance, using ATPase and photoaffinity displacement assays. Key reactions include a new alkoxide-mesylate displacement for the assembly of the PEG linkers and a double Sonogashira coupling reaction.

Full text of DOI:10.1016/S0040-4039(01)00619-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 3819–3822
Synthesis and analysis of polyethylene glycol linked
P-glycoprotein-specific homodimers based on (−)-stipiamide
Merritt B. Andrus,a,* Timothy M. Turner,a Emily P. Updegraff,a Zuben E. Saunab and
Suresh V. Ambudkarb
aBrigham Young University, Department of Chemistry and Biochemistry, C100 BNSN, Provo, UT 84602-5700, USA
bNational Cancer Institute, NIH, Laboratory of Cell Biology, Division of Basic Sciences, Building 37,
Room 1B-17 37 Convent Dr., MSC4255 Bethesda, MD 20878-4255, USA
Received 27 November 2000; accepted 11 April 2001
Abstract—A series of five homodimeric polyethylene glycol (PEG) linked homodimers based on the multidrug resistance reversal
agent (−)-stipiamide were made and tested for their ability to interact with P-glycoprotein, the protein responsible for multidrug
resistance, using ATPase and photoaffinity displacement assays. Key reactions include a new alkoxide-mesylate displacement for
the assembly of the PEG linkers and a double Sonogashira coupling reaction. © 2001 Elsevier Science Ltd. All rights reserved.
Polyvalency has become
a
popular strategy for
compatibility with assays. A particularly challenging
target is P-glycoprotein (Pgp) the ATP-dependent
drug efflux pump whose overexpression confers multi-
drug resistance (MDR) to cancer cells. The develop-
ment of resistance in cancer cells to chemotherapeutic
agents has been a major impediment to effective clini-
cal treatments.7 Recently we reported synthetic routes
to the new MDR reversal agents, (−)-stipiamide 1, a
highly toxic polyene and a designed, more potent,
nontoxic compound 6,7-dehydrostipiamide (DHS) that
restores the cytotoxicity of doxorubicin to resistant
human breast cancer cells (MCF-7adrR) at low con-
centration (Scheme 1).8 A solution-phase combinato-
rial library of DHS compounds was generated using a
key Sonogashira coupling reaction and was screened.9
increasing the binding affinity of ligands to multi-
meric receptors.1 A weak binder can be converted
into a polyvalent tight binder provided the receptor
possesses various binding sites or can be dimerized.
The origin of the overall enhanced binding can be
attributed to the more favorable second and subse-
quent binding events where much less entropy is lost
due to induced proximity. Among the successful
examples are the FK506 and cyclosporin dimers of
Schreiber,2 the vancomycin dimers of Whitesides3 and
Griffin,4 and the polysaccharides of Kiessling,5
Whitesides, and Fan.6 Critical issues include the
nature of the linker and the polymeric support, the
length and position of attachment of the linker, and
Ph
Me
truncated-DHS 2
(-)-stipiamide 1
Me
Me
OH
Me Me
Me
O
2-nap
Me
O
Me
NH
N
HO
OH
Me
O
Me Me
Me Me
OH
Me
2-nap
2-nap
Me
H
N
H
N
OH
O
O
O
X
Dual Domain Stipiamide Homodimers (X= 0, 2, 5, 8,12) DDSH 3
Scheme 1.
Keywords: etherification; polyethylene glycol; dimerization; anticancer compounds; multidrug resistance.
* Corresponding author.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00619-0

3820
M. B. Andrus et al. / Tetrahedron Letters 42 (2001) 3819–3822
Many potent, non-natural compounds were identified
including the 2-naphthyl morpholino amide 2 shown.
polyethylene glycols that are not commercially avail-
able.13 Unlike previous reports that have employed
chlorides14 and tosylates15 as leaving groups, we have
found that mesylates function as the most efficient
We now report initial efforts to further increase MDR
reversal potency by creating bivalent homodimeric
polyenes based on stipiamide linked with ethylene gly-
col spacers 3 (Scheme 1). Photoaffinity labeling has
been used to identify two non-identical sites most likely
formed by the TM5-6 and TM11-12 regions of Pgp.10,11
Close proximity of the cytosolic ATP-binding sites sup-
ports the notion that ATP hydrolysis induces confor-
mational changes that are conveyed to the TM regions
leading to drug displacement and efflux. Conforma-
tional changes in the substrate-binding domain follow-
ing ATP hydrolysis have recently been determined.11
coupling partners. Diols
4 (x=2, 4) were first
monobenzylated using 50% aqueous hydroxide at reflux
to give the protected alcohols 5 (Scheme 2). The key
step that allowed for reproducible glycol production
employed sodium hydride with alcohol 5 followed by
dropwise addition of dimesylate 616 and reflux to
provide 7.17 Hydrogenation (200 psi) with palladium on
carbon (10%) in methanol was used to give diols 8.
Dimesylate formation and sodium azide displacement
provided diazides 9. Commercially available hexa-
ethylene glycol 8 (x=5) was used for conversion to 9
(x=5). Reduction with triphenylphosphine was used to
generate the key diamines 10. Completing the series are
10 (x=0 and 2) which were purchased.
,
Additionally, a low resolution (25 A) electron diffrac-
tion structure for Pgp shows the protein to be doughnut
12
,
shaped with an approximate 50 A pore opening. The
proximity of the TM5-6 and TM11-12 helices and the
identity of residues involved in substrate binding
remain unknown. Success of the polyvalency approach
requires a proper linker length in order to span the
binding sites. A new approach involving displacement
of a dimesylate was used to access PEG linkers that
Treatment of the diamines 10 (1.2 equiv.) at −40°C with
the known acid (1 equiv.) generated from carboxylic
acid 119b led to the formation of the diamides 12
(Scheme 3). Difficulties with low yields in some cases
(x=8, 12) may be attributed to the very hydroscopic
nature of the diamine glycols. Double Sonogashira
couplings were then performed with the bis-amide alky-
nes 12 reacted with the previously reported vinyl iodide
139b (3 equiv.) under palladium chloride, copper iodide
catalysis. These optimized conditions include diiso-
propylamine together with the moderately polar ethyl
acetate as solvent and a starting temperature of −20°C
followed by immediate warming to room tempera-
,
now provide for a distance range from 3 to 50 A.
Double Sonogashira couplings, using a recently devel-
oped set of conditions, were also used to generate the
homodimers 3. Pgp inhibition and binding was deter-
mined using ATPase and competitive displacement of
photoaffinity analog binding assays.
The key to the synthesis of the homodimers was the
development of the general, efficient route to the
OBn
BnO
OMs
NaH, THF
BnCl,
NaOH
OH
BnO
OH
HO
O
O
O
MsO
X
O
2
X
∆
X
7 X= 8 (73%)
5 X= 2 (73%)
6
4 X= 2, 4
X= 12 (59%)
X= 4 (81%)
1) MsCl,
Et3N
2) NaN3,
EtOH
NH2
H2N
N3
OH
N3
HO
O
H2, Pd/C 10%
MeOH
O
PPh3, THF, H2O
O
X
X
X
10 X= 0 and 2 (Aldrich)
X= 5 (83%)
8 X= 5 (Aldrich)
X= 8 (88%)
X= 12 (77%)
9 X= 5 (63%)
X= 8 (95%)
X= 12 (88%)
X= 8 (95%)
X= 12 (89%)
Scheme 2.
Me
1) (COCl)2, DMF
Me
H
H
N
Me
2)
N
NH2
H2N
OH
O
O
X
O
X
10
O
O
12 X= 0 (74%)
X= 2 (39%)
11
X= 5 (80%)
X= 8 (44%)
Me Me
X= 12 (25%)
(PH3P)2PdCl2
CuI, i-Pr2NH,
EtOAc
2-nap
I
13
OH
Me Me
OH
Me Me
OH
2-nap
Me
O
Me
2-nap
H
N
H
N
O
O
X
DDSH 3 X= 0 (55%), X= 2 (53%), X= 5 (53%), X= 8 (76%), X= 12 (71%)
Scheme 3.

M. B. Andrus et al. / Tetrahedron Letters 42 (2001) 3819–3822
3821
1) TBSCl,
Et3N 76%
2) MsCl,
Et3N
3) NaN3,
EtOH
1) PPh3 78%
2)
Me
Me
OH
HO
H
N
O
OH
Cl
OTBS
N3
O
O
X
Et3N 69%
3) TBAF
O
79%
O
X
X
8 X= 5
15 X= 5
14 X= 5
71%
Me Me
Me Me
OH
2-nap
2-nap
Me
(PH3P)2PdCl2
CuI, i-Pr2NH,
EtOAc
I
H
N
OH
13
OH
O
O
X
16 X= 5
41%
Scheme 4.
ture.8b The yields for this double process were moderate
(50%) for the shorted linked amides 3 (x=0, 2, 5) and
good (70%) for the longer products 3 (x=8, 12).18
These conditions were developed previously for the
synthesis of the stipiamide variants and were also used
in the solution-phase library.9 More polar solvents
DMF and NMP, and less polar solvents THF and
benzene, gave much lower yields at near 1:1 stoichiome-
try. Beginning the reaction at rt also lowered product
yields. Previous Sonogashira coupling examples have
shown that amides containing acetylenes are very poor
coupling partners.19 Recently this new Sonogashira pro-
tocol has been employed as the key step in the synthesis
of callipeltoside.20
Table 1. ATPase activity, and inhibition of IAAP binding
by homodimers 3 and control amide 16
a
,
Cmpd
x
d (A)
ATPase activityb
Kic
2
16
3
3
3
3.20
0.34
3.30
1.90
1.90
0.70
0.00
2.7
7.6
19.1
9.7
15.1
10.1
1.7
0
2
5
8
12
3
11
22
35
50
3
3
a Approximate distance between amide nitrogens.
b ATP-hydrolysis, fraction of control, values >1 represent stimula-
tion, values <1 inhibition.
c Ki (mM) inhibition of 125IAAP binding to Pgp.
A monomeric ethylene glycol-amide reversal agent was
also made for control purposes in the MDR and Pgp
assays (Scheme 4). Monosilyl ether protection, mesyla-
tion and azide displacement were uneventful with hexa-
ethylene glycol 8 (x=5) to give 14. Phosphine reduc-
tion, acid chloride coupling, and TBAF deportation
generated amide 15. Coupling with 13 then gave the
desired control substrate 16.
dimers that varies greatly with tether length. As the
tether length increases, ATPase activity goes from being
stimulatory to strongly inhibitory. Control 16, contain-
ing only a PEG tail also inhibits ATPase activity, but
not to the same degree as 3 (x=12). Also the Ki of 16
is less potent at 7.6 mM. This suggests that there may be
potential to reverse MDR using a dimer strategy.
Assays using drug resistant MCF7adrR and NIH-3T3
cells, to be reported elsewhere, will be used to establish
reversal concentrations for the dimers along with more
detailed Pgp binding studies to further correlate tether
length. These methods and results will be used to
develop other polyvalent agents to achieve stronger
interaction with Pgp and more effective reversal of
MDR.
Preliminary activity of the dimers along with the
monomeric control agents 2 and 16 were investigated
using Pgp ATPase activity and photoaffinity displace-
ment assays (Table 1). Stimulation of Pgp ATPase
activity was determined along with displacement of the
prazosin photoaffinity label, iodoarylazidoprazosin
(
125IAAP) following the known protocol.21 The dimer
effect on both ATPase activity and IAAP binding to
Pgp appear to be significant. ATPase stimulation
reaches a maximum at low concentration (1 mM) and
steadily drops off as the concentration is increased (not
shown). In contrast 3 (x=5) slowly achieves maximum
stimulation up to 50 mM and this level is maintained as
the concentration increases. The other dimers are simi-
lar to other known monomeric MDR reversal com-
pounds where maximum stimulation is achieved and
rapidly drops off as concentration increases. Most
remarkable is the dimer 3 (x=12) where near complete
ATPase inhibition was found. Potent interaction with
Pgp is also seen in the photoaffinity displacement
results where 3 (x=12) binds most tightly at 1.7 mM.
The other dimers were less effective at 125IAAP dis-
placement with Kis in the 10–20 mM range. These
results clearly indicate an interaction with Pgp for the
Acknowledgements
We wish to thank the National Institutes of Health
(GM57275 to MBA), The American Cancer Society,
and The Brigham Young University Cancer Center for
funding, and Dr. Bruce Jackson for mass spectrometry,
and Vicki Croy (Purdue U.) for the MDR assays.
References
1. Mammen, M.; Choi, S.-K.; Whitesides, G. M. Angew.
Chem., Int. Ed. 1998, 37, 2754.

3822
M. B. Andrus et al. / Tetrahedron Letters 42 (2001) 3819–3822
2. Crabtree, G. R.; Schreiber, S. L. Trends Biochem. Sci.
filtered, and concentrated to give a yellow oil. Flash
chromatography (100% EtOAc to 10% MeOH/EtOAc)
gave the desired product as an oil: 4.51 g (73% yield).
Rf=0.34 (10% MeOH/EtOAc); 1H NMR (300 MHz,
CDCl3) l 7.34–7.26 (m, 10H), 4.56 (s, 4H), 3.69–3.61 (m,
36H); 13C NMR (75 MHz, CDCl3) l 138.4, 128.5, 127.9,
127.8, 73.4, 70.8, 70.7, 69.6; HRMS FAB (M+Na) calcd
for C32H50O10Na 617.3287, found 617.3272.
1996, 21, 418.
3. Jianghong, R.; Yan, L.; Lahiri, J.; Whitesides, G. M.;
Weis, R. M.; Warren, H. S. Chem. Biol. 1999, 6, 353.
4. Sundram, U. N.; Griffin, J. H.; Nicas, T. I. J. Am. Chem.
Soc. 1996, 118, 13107.
5. Kanai, M.; Mortell, K. H.; Kiessling, L. L. J. Am. Chem.
Soc. 1997, 119, 9931.
6. Mammen, M.; Helmerson, K.; Kishore, R.; Choi, S.-K.;
Phillips, W. D.; Whitesides, G. M. Chem. Biol. 1996, 3,
757.
7. (a) Ambudkar, S. V.; Dey, S.; Hrycyna, C. A.;
Ramachandra, M.; Pastan, I.; Gottesman, M. M. Annu.
Rev. Pharmacol. Toxicol. 1999, 39, 361; (b) Gottesman,
M. M.; Pastan, I. Annu. Rev. Biochem. 1993, 62, 385.
8. (a) Andrus, M. B.; Lepore, S. D. J. Am. Chem. Soc. 1997,
119, 2327; (b) Andrus, M. B.; Lepore, S. D.; Turner, T.
M. J. Am. Chem. Soc. 1997, 119, 12159.
Preparation of 7 (x=12): Same conditions as above
except that 5 (x=4) was used (59% yield). Rf=0.21 (10%
MeOH/EtOAc); 1H NMR (300 MHz, CDCl3) l 7.35–
7.27 (m, 10H), 4.57 (s, 4H), 3.69–3.60 (m, 52H); 13C
NMR (75 MHz, CDCl3) l 138.4, 128.5, 127.9, 127.8,
73.4, 70.8, 70.7, 69.6; HRMS FAB (M+Na) calcd for
C40H66O14Na 793.4362, found 793.4374.
18. Preparation of 3: To a flask containing dialkynylamide 12
(x=0) (11.5 mg, 0.038 mmol) was added 13 (47 mg, 0.115
mmol) and EtOAc (3.9 mL). The reaction was cooled to
−20°C and (Ph3P)2PdCl2 (4.1 mg, 0.006 mmol), CuI (3.8
mg, 0.02 mmol), and i-Pr2NH (0.58 mL, 0.2 M) were
added. The flask was removed from the cold bath and
allowed to warm to rt. The solution was filtered through
a silica gel plug using 30% MeOH/EtOAc and concen-
trated. Purification via radial chromatography (100%
EtOAc) gave 18.0 mg (55% yield) of 3 (x=0) as an oil.
Rf=0.5 (2% MeOH/EtOAc); 1H NMR (300 MHz,
CDCl3) l 7.81–7.67 (m, 6H), 7.60 (s, 2H), 7.47–7.40 (m,
4H), 7.34 (dd, J=8.4, 1.5 Hz, 2H), 6.61 (bs, 2H), 6.38 (t,
J=5.7 Hz, 2H), 5.98 (dd, J=15.9, 8.4 Hz, 2H), 5.52 (dd,
J=15.9, 0.9 Hz, 2H), 5.45 (t, J=6.6 Hz, 2H), 3.63 (d,
J=8.7 Hz, 2H), 3.48–3.46 (m, 4H), 2.87–2.80 (m, 4H),
2.49–2.29 (m, 14H), 1.85 (s, 2H), 1.84 (s, 6H), 1.56 (s,
6H), 0.77 (d, J=6.6 Hz, 6H); 13C NMR (75.5 MHz,
CDCl3) l 170.6, 145.9, 139.6, 135.9, 134.6, 133.7, 132.1,
131.9, 128.3, 128.0, 127.8, 127.6, 127.5, 126.6, 126.1,
125.3, 111.3, 88.4, 81.8, 79.9, 41.5, 40.7, 35.9, 29.6, 28.0,
19.1, 16.9, 13.1, 11.2; HRMS FAB (M+Na) calcd for
C58H68O4N2Na 879.5071, found 879.5065. Compounds 3
(x=2, 5, 8, 12) were formed followed the same procedure
with yields indicated (Scheme 3).
19. (a) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetra-
hedron Lett. 1975, 4467; (b) Rossi, R.; Carpita, A.; Cio-
falo, M.; Lippolis, V. Tetrahedron 1991, 47, 8443; (c)
Rossi, R.; Carpita, A.; Bellina, F. Org. React. Proc. 1995,
129.
20. Paterson, I.; Davies, R. D. M.; Marquez, R. Angew.
Chem., Int. Ed. 2001, 40, 603.
21. (a) Ambudkar, S. V. Methods Enzymol. 1998, 292, 504–
514; (b) Dey, S.; Ramachandra, M.; Pastan, I.; Gottes-
man, M. M.; Ambudkar, S. V. Methods Enzymol. 1998,
292, 318.
9. (a) Andrus, M. B.; Turner, T. M.; Asgari, D.; Wenke, L.
J. Org. Chem. 1999, 64, 2978; (b) Andrus, M. B.; Turner,
T. M.; Sauna, Z. E.; Ambudkar, S. V. J. Org. Chem.
2000, 65, 4973–4983.
10. (a) Germann, U. A. Eur. J. Cancer 1996, 32A, 927; (b)
Greenberger, L. M. J. Biol. Chem. 1993, 268, 11417.
11. (a) Ramachandra, M.; Ambudkar, S. V.; Chen, D.;
Hrycyna, C. A.; Dey, S.; Gottesman, M. M.; Pastan, I.
Biochemistry 1998, 37, 5010; (b) Dey, S.; Ramachandra,
M.; Pastan, I.; Gottesman, M. M.; Ambudkar, S. V.
Proc. Natl. Acad. Sci. 1997, 94, 10594; (c) Sauna, Z. E.;
Ambudkar, S. V. Proc. Natl. Acad. Sci. 2000, 97, 2515.
12. Rosenberg, M. F.; Callaghan, R.; Ford, R. C.; Higgins,
C. F. J. Biol. Chem. 1997, 272, 10685.
13. Zalipsky, S. Bioconjugate Chem. 1995, 6, 150.
14. Coudert, G.; Mpassi, M.; Guillaumet, G.; Selve, C.
Synth. Commun. 1986, 16, 19.
15. Keegstra, E. M. D.; Zwikker, J. W.; Roest, M. R.;
Jenneskens, L. W. J. Org. Chem. 1992, 57, 6678.
16. All intermediates were characterized by 1H and 13C
NMR, and HRMS. Dimesylate 6 was produced from
triethylene glycol, mesyl chloride (2.1 equiv.), and triethyl-
amide (2.4 equiv.) in methylene chloride (0.2 M). Sodium
bicarbonate work-up and silica gel chromatography were
used to isolate the product (97%).
17. Preparation of 7 (x=8): 95% NaH (0.526 g, 20 mmol)
was added to 20 mL THF followed by tri(ethylene glycol)
benzylether 5 (x=2) (5.0 g, 20 mmol in 12 mL THF) and
the mixture was stirred for 1 h. Tri(ethylene glycol)
dimesylate 6 (3.18 g, 10 mmol in 7.5 mL THF) was added
dropwise over 30 min and the solution was refluxed for
20 h. Upon cooling, the solution was diluted with
NaHCO3 (satd) (250 mL) and extracted with CH2Cl2.
The combined organic layers were dried (MgSO4),
.