Synthesis and evaluation of antiproliferative activity of a geldanamycin-Herceptin(TM) immunoconjugate

Article abstract of DOI:10.1016/S0960-894X(00)00155-4

Geldanamycin was modified with 1,4-diaminobutane to introduce a primary amine that was subsequently employed to provide a maleimide for protein linkage. Monoclonal antibody Herceptin(TM) was then derivatized to generate thiol groups that reacted with the maleimide derivative to produce the immunoconjugate. The product showed antiproliferative activity greater than native Herceptin(TM). (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00155-4

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1025±1028
Synthesis and Evaluation of Antiproliferative Activity of
a Geldanamycin±HerceptinTM Immunoconjugate
Raya Mandler, a Ekaterina Dadachova, b James K. Brechbiel, b
Thomas A. Waldmann a and Martin W. Brechbiel b,*
aMetabolism Branch, National Cancer Institute, NIH, Bethesda, MD 20982-1002, USA
bRadioimmune & Inorganic Chemistry Section, Radiation Oncology Branch, National Cancer Institute,
NIH, Bethesda, MD 20982-1002, USA
Received 28 October 1999; accepted 1 March 2000
AbstractÐGeldanamycin was modi®ed with 1,4-diaminobutane to introduce a primary amine that was subsequently employed to
provide a maleimide for protein linkage. Monoclonal antibody HerceptinTM was then derivatized to generate thiol groups that
reacted with the maleimide derivative to produce the immunoconjugate. The product showed antiproliferative activity greater than
native HerceptinTM. # 2000 Elsevier Science Ltd. All rights reserved.
The advent of monoclonal antibody (mAb) technology
raised hopes that cancer therapy would become a reality
with the ability to develop high speci®city agents that
could target tumor-associated markers.1 However, the
clinical trials with unconjugated, or unmodi®ed mAbs
have generally yielded only partial responses because
most mAbs possess minimal antitumor activity.2 To
compensate, toxins have been fused with, or drugs have
been conjugated to mAbs. Trials with such immuno-
conjugates resulted in promising outcomes but have also
been fraught with adverse complications.3 Currently, one
of the most promising targets for immunotherapy is the
membrane receptor HER2, a member of the epidermal
growth factor receptor family. While only marginally
detected in adult tissues,4,5 it is over-expressed in
approximately 30% of human gastric, lung, and breast
carcinomas.4,6 8 HER2 over-expression in breast carcino-
mas is inversely related to estrogen receptor expression
and is correlated with poor prognosis.9
when Herceptin is the sole therapeutic agent. We,
therefore, set out to augment its activity by forming an
immunoconjugate with the highly cytotoxic drug geldan-
amycin (GA).
GA (Fig. 1) is a benzoquinoid ansamycin produced by
the actinomycete streptomyces hygroscopicus and is
related to herbimycin A.16 It binds with high anity to
the cytosolic protein chaperone hsp90 and disrupts its
ability to protect cellular enzymes from proteasomal
degradation.17 Since HER2 stability depends on inter-
action with hsp90, GA causes accelerated elimination of
this receptor.18 20 and disruption of the proper proces-
sing of newly synthesized HER2 molecules into the
endoplasmic reticulum.21
While the antitumor potential of GA has long been
recognized, clinical use was not pursued due to its severe
toxicity and diculties with aqueous formulation. Thus,
numerous derivatives have been synthesized and studied
in an e€ort to develop more selective agents with anti-
tumor activity.22 These studies indicated that modest
modi®cations at the 17-position on the quinone ring
maintain cytotoxicity at nanomolar range. Although
GA itself cannot be directly linked to proteins, after
introduction of a primary amine group at the 17-posi-
tion, a hetero-bifunctional cross-linking reagent could
be employed to conjugate GA to mAbs. One such deri-
vative, 17-(3-aminopropylamino)-GA (17-APA-GA)
with a maleimide linker added (17-GMB-APA-GA)
(Fig. 1), has been conjugated to the anti-HER2 mAb
Blocking HER2 activity or interfering with its expres-
sion has been shown to inhibit proliferation and reduce
tumor growth.4,6,10 Thus, several anti-HER2 mAbs have
entered clinical trials in the past 6 years,11 15 and the
humanized anti-HER2 antibody HerceptinTM is currently
used clinically in metastatic breast cancer therapy.11,13
However, the objective response rate is relatively low
*Corresponding author. Fax: +1-301-402-1923; e-mail: martinwb@
box-m.nih.gov
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00155-4

1026
R. Mandler et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1025±1028
Figure 1.
e21 and its ability to augment both the anti-proliferative
e€ect of e21 and its elimination of HER2 has been
demonstrated.23 However, preparation of this derivative
has been problematic. Preparation of 17-APA-GA, as
described in the literature,22 produced a complex mix-
ture of products.24 Exposure to silica catalysed forma-
tion of a seven-membered imine ring from the distal
amine and the quinone carbonyl precluding use of this
medium for isolation of the desired primary amine.
Employment of other chromatographic media has not
yielded satisfactory results.24 While milligram amounts
of 17-APA-GA have been successfully obtained, the
immunoconjugate formed with this derivative contained
higher molecular weight by-products necessitating pur-
i®cation of the immunoconjugate by preparative HPLC
to ensure purity (vida infra).
be 1±1.5 equiv of diamine and 3±5 days of reaction time
while carefully following the progress of the reaction by
TLC.25 Use of larger excesses of diamine tended to sig-
ni®cantly increase production of a compound that
appeared to have added 2 equiv of diamine as determined
by mass spectrometry, presumably due to formation of
an imine on the quinone.
The 17-ABA-GA was then treated with the hetero-
bifunctional cross-linking reagent GMB to introduce a
maleimide.26 Isolation of this product by column chro-
matography was uncomplicated and produced a reagent
suitable for conjugation to a monoclonal antibody, 17-
GMB-ABA-GA (73%).
In order to perform the conjugation, 17-GMB-ABA-
GA was dissolved in DMSO (2 mg/mL) prior to addi-
tion to the conjugation reaction. Thiolation of the mAb
with Traut's reagent was carried out as previously descri-
bed26 and excess reagent was removed by bu€er exchange
into conjugation bu€er (50 mM HEPES, 150 mM NaCl,
10 mM EDTA pH 7.0). The mAb was reacted with 17-
GMB-ABA-GA and the reaction mixture was kept in the
dark at 25ꢀC for 1 h and then dialyzed extensively (1 LÂ3
during 48 h) against PBS without Ca+2/Mg+2 at 4 ꢀC.
The presence of the GA moiety on the mAb was con-
Published data suggested that the 17-(4-aminobutyl-
amino)-GA (17-ABA-GA) (Fig. 2) might be a better
derivative for the production of adequate amounts of a
biologically active GA immunoconjugate. The biological
activity of 17-ABA-GA was expected to be mildly lower
than that of 17-APA-GA, if one extrapolates from the
IC50 data against SKBr-3 cells for 17-ABA-GA, 17-N,N-
dimethylABA-GA, and similar chain length deriva-
tives.22 However, the synthesis and isolation of 17-ABA-
GA was not expected to be hindered by imine cyclization
to form an eight-membered ring.
®rmed by spectrophotometric reading at A334
.
The purity of the immunoconjugate was found to be
signi®cantly improved compared with that of 17-GMB-
APA-GA. The analytical size-exclusion HPLC27 chro-
matograms clearly indicated that the 17-GMB-ABA-GA
immunoconjugate was essentially congruent with Her-
ceptinTM, while the previously prepared analogous 17-
APA-GA immunoconjugate contained signi®cant levels
of by-products (Fig. 3). These results were con®rmed by
SDS-PAGE as well (data not shown).
Treatment of GA with 1,4-diaminobutane (Fig. 2) pro-
duced a dark-purple solid that was isolated by pre-
cipitation with hexane.25 Thin-layer chromatography
indicated several products and after careful column
chromatography on silica gel, 17-ABA-GA was success-
fully isolated in 84% yield. The product distribution was
highly dependent upon both reaction time and the addi-
tion of excess diamine. Optimal conditions appeared to
Figure 2.

R. Mandler et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1025±1028
1027
Figure 3.
Figure 4.
The antiproliferative action of the immunoconjugate
thus formed was evaluated by testing its activity on
MDA-361/DYT2, a HER2-positive human breast car-
cinoma cell line.28 The cells were seeded in 96-well, ¯at-
bottom plates and allowed to adhere. HerceptinTM, in
either free or conjugated form, was added to the wells in
serial dilutions (range of 1.5 to 0.01 mg/mL). The cells
were allowed to proliferate until the untreated control
cultures reached 80% con¯uency (usually 5 days). Cell
proliferation was assessed by the crystal violet staining
method after washing the cultures in PBS and ®xing the
cells with 80% ethanol.29 As depicted in Figure 4, Her-
ceptinTM had minimal antiproliferative activity, redu-
cing cellular growth by only 30% in the HER2-positive
cell line at 1.5 mg/mL.30 In contrast, at the same con-
centration, the HerceptinTM 17-ABA-GA conjugate,
H:ABA-GA, reduced proliferation by 85%. The IC50 of
the immunoconjugate was determined from four separate
experiments to be 0.5Æ0.03 mg/mL. In contrast, IC50
values for native HerceptinTM could not be determined
as in the native form HerceptinTM maximal inhibition
reached only 35%. Furthermore, the immunoconjugate
did not inhibit the HER2 negative cells (Fig. 4).
cultures were 1.4, 1.6 and 6.2 days, respectively. Thus,
H:ABA-GA caused approximately a 4-fold reduction in
growth as compared to unmodi®ed HerceptinTM
.
In this study we have shown that conjugation of the
cytotoxic derivative of GA, 17-ABA-GA, to an anti-
HER2 mAb was successful in retaining both the anti-
proliferative activity and the speci®city towards HER2-
expressing neoplastic cells. Furthermore, the immuno-
conjugate exhibited enhanced antiproliferative action
when compared with that of the native mAb. Addition-
ally, we have demonstrated that the insertion of a single
methylene in the GA derivative signi®cantly reduced the
technical obstacles encountered in larger-scale produc-
tion of a 17-APA-GA immunoconjugates. Ongoing
biological experiments are under way to evaluate in vivo
ecacy of these conjugates in animal xenograft model.
Acknowledgements
We thank Gordon Cragg, NPG, DTP, DCTD, NCI,
NIH for his very generous supply of geldanamycin,
Robert Cohen and Mark Sliwkowski, Genentech Inc.,
for their generous supply of HerceptinTM, and the
Laboratory of Analytical Chemistry, NIDDK, for their
assistance in obtaining Mass Spectra data for the gel-
danamycin derivatives.
In addition, the cell doubling time was measured in
MDA-361/DYT2 cultures treated with 0.5 mg/mL Her-
ceptinTM or H:ABA-GA. Cell doubling time for the
control (PBS), HerceptinTM, or H:ABA-GA treated

1028
R. Mandler et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1025±1028
(66 mg, 0.75 mmol) was added. The reaction was left stirring
References and Notes
shielded from light for at least 72 h during which the golden
yellow of the solution was replaced with a deep purple. The
reaction was monitored by TLC on silica (10% MeOH:
CHCl3, Rf=0.03 and 10% MeOH:CHCl3:1% Et3N, Rf=0.2±
0.34). Additional diamine was added as needed and the reac-
tion permitted to continue until TLC indicated near complete
consumption of the GA. The solvent was removed with mini-
mal heating to near dryness and the crude product precipitated
with addition of ꢁ1 L of hexane. The product was collected
and dried. The 17-ABA-GA was isolated by ¯ash chromato-
graphy on silica (2.5Â35 cm column). Trace amounts of GA
and several unidenti®ed side products were eluted with
increasing percentages of MeOH in CHCl3 up to 10%
MeOH:CHCl3. Addition of 1% Et3N to that solvent then
permitted clean elution of the product while subsequent
increase of MeOH to 20% permitted elution of the presumed
bis-amine adduct as determined by CI±MS. 1H NMR (300
MHz, CDCl3) d 0.899 (d, 3H, J=6.9 Hz), 0.936 (d, 3H, J=6.9
Hz), 1.504 (m, 1H), 1.667 (m, 2H), 1.735 (s, 3H), 1.960 (s, 3H),
2.356 (dd, 1H, J=13.2, 9.9 Hz), 2.57.77 (m, 9H), 3.206 (s, 3H),
3.301 (s, 3H), 3.405 (m, 3H), 3.514 (m, 3H), 4.248 (d, 1H,
J=9.6 Hz), 4.769 (br. S, 2H), 5.124 (s, 1H), 5.75.90 (m, 2H),
6.380 (t, 1H, J=5.7 Hz), 6.523 (t, 1H, J=14.4 Hz), 6.895 d,
1H, J=11.7 Hz), 7.204 (s, 1H), 9.127 (s, 1H); CI±MS or FAB±
MS (NH3 or nba) m/e 617 (M++1). Anal. calcd for
C32H48N4O8: C, 62.30; H, 7.86; N, 9.09; found: C, 62.42; H,
8.04; N, 9.32. The 17-ABA-GA (250 mg, 0.406 mmol) was
dissolved in EtOAc (30 mL) and treated with GMB (133 mg,
0.45 mmol). TLC on silica with 10% MeOH:CHCl3 indicated
one major component, (10% MeOH:CHCl3, Rf ˆ 0:32 and
10% MeOH:CHCl3:1% Et3N, Rf ˆ 0:73). The solvent was
removed without heating and the product isolated by ¯ash
chromatography on silica as above eluting with 2% MeOH/
CHCl3 (230 mg). 1H NMR (300 MHz, CDCl3) d 0.900 (d, 3H,
J=6.6 Hz), 0.933 (d, 3H, J=6.9 Hz), 1.605 (m, 1H), 1.657 (m,
2H), 1.732 (s, 3H), 1.872 (m, 4H), 1.963 (s, 3H), 2.096 (t, 2H,
J=6.6 Hz), 2.305 (t, 2H, J=6.9 Hz), 2.50.78 (m, 4H), 3.206 (s,
3H), 3.264 (q, 2H, J=6.9 Hz), 3.301 (s, 3H), 3.35.60 (m, 8H),
4.248 (d, 1H, J=9.6 Hz), 5.125 (s, 1H), 5.75.90 (m, 2H), 5.944
(t, 1H, J=6.0 Hz), 6.234 (t, 1H, J=5.1 Hz), 6.523 (t, 1H,
J=11.7 Hz), 6.644 (s, 2H), 6.8992 (d, 1H, J=11.7 Hz), 7.204
(s, 1H), 9.114 (s, 1H); CI±MS/FAB±MS (NH3/nba) m/e 782
(M++1); HR±FAB±MS m/e=913.3146 error= 1.8 ppm.
Anal. calcd for C40H55N5O11: C, 61.43; H, 7.10; N, 8.96;
found: C, 61.57; H, 7.23; N, 8.69.
1. Kohler, G.; Milstein, C. Nature 1975, 256, 495.
2. Grossbard, M. L.; Press, O. W.; Appelbaum, F. R.; Bern-
stein, I. D.; Nadler, L. M. Blood 1992, 80, 863.
3. King, D. J. In Applications and Engineering of Monoclonal
Antibodies; Taylor & Francis: Philadelphia, PA; Chapter 4.
4. Hynes, N. E.; Stern, D. F. Biochem. Biophys. Acta 1994,
1198, 165.
5. Lee, K. F.; Simon, H.; Chen, H.; Bates, B.; Hung, M. C.;
Hauser, C. Nature 1995, 378, 394.
6. Revillion, F.; Bonneterre, J.; Peyrat, J. P. Eur. J. Cancer
1998, 34, 791.
7. Slamon, D. J.; Clark, G. M.; Wong, S. G.; Levin, W. J.;
Ullrich, A.; McGuire, W. L. Science 1987, 235, 177.
8. Iglehart, J. D.; Kraus, M. H.; Langton, B. C.; Huper, G.;
Kerns, B. J.; Marks, J. R. Cancer Res. 1990, 50, 6701.
9. Muss, H. B.; Thor, A. D.; Berry, D. A.; Kute, T.; Liu, E. T.;
Koerner, F.; Cirrincione, C. T.; Budman, D. R.; Wood, W. C.;
Barcos, M.; Hendrson, I. C. New Eng. J. Med. 1994, 330,
1260.
10. Hynes, N. E. Sem. Cancer Biol. 1993, 4, 19.
11. Baselga, J.; Tripathy, D.; Mendelsohn, J.; Baughman, S.;
Benz, C. C.; Dantis, L.; Sklarin, N. T.; Seidman, A. D.; Hudis,
C. A.; Moore, J.; Rosen, P. P.; Twaddell, T.; Henderson, I. C.
Norton, L. J. Clin. Oncol. 1996, 14, 737.
12. Pegram, M. D.; Baly, D.; Wirth, C.; Gilkerson, E.; Sla-
mon, D. J.; Sliwkowski, M. K. Proc. Am. Assoc. Cancer Res.
1997, 38, 602.
13. Clark, J. I.; Alpaugh, R. K.; von Mehren, M.; Schultz, J.;
Gralow, J. R.; Cheever, M. A.; Ring, D. B.; Weiner, L. M.
Cancer Immunol. Immunotherap. 1997, 44, 265.
14. Curnow, R. T. Cancer Immunol. Immunotherap. 1997, 45,
210.
15. Hung, M. C.; Chang, J. Y. J.; Xing, X. M. Adv. Drug
Delivery Rev. 1998, 30, 219.
16. DeBoer, C.; Meulman, P. A.; Wnuk, R. J.; Peterson, D. H.
J. Antibiotics 1970, 23, 442.
17. Lawson, B.; Brewer, J. W.; Hendershot, L. M. J. Cell
Physiol. 1998, 174, 170.
18. Toft, D. O. TEM 1998, 9, 238.
19. Scheibel, T.; Buchner, J. Biochem. Pharmacol. 1998, 56,
675.
20. Hartmann, F.; Horak, E. M.; Cho, C.; Lupu, R.; Bolen, J.
B.; Stetler-Stevenson, M. A.; Pfreundschuh, M.; Waldmann,
T. A.; Horak, I. D. Int. J. Cancer 1997, 70, 221.
21. Mimnaugh, E. G.; Chavany, C.; Neckers, L. J. Biol. Chem.
1996, 271, 22796.
26. Wu, C.; Brechbiel, M. W.; Kozak, R. W.; Gansow, O. A.
Bioorg. Chem. Med. Lett. 1994, 4, 449.
22. Schnur, R. C.; Corman, M. L.; Gallaschun, R. J.; Cooper,
B. A.; Dee, M. F.; Doty, J. L.; Muzzi, M. L.; Moyer, J. D.;
DiOrio, C. I.; Barbacci, E. G.; Miller, P. E.; O'Brien, A. T.;
Motin, M. J.; Foster, B. A.; Polack, V. A.; Savage, D. M.;
Sloan, D. E.; Pustilnik, L. R.; Moyer, M. P. J. Med. Chem.
1995, 38, 3806.
23. Mandler, R.; Wu, C.; Brechbiel, M.W.; Roettinger, A. J.;
Ho, D. K.; Newman, D. J.; Sausville, E. A.; King, C. R.;
Yang, D.; Lippman, M. E.; Waldmann, T. A. J. Natl. Cancer
Inst. submitted for publication.
27. HPLC was performed using a Dionex system equipped
with a Model GPM-2 gradient pump, Waters 717plus Auto-
sampler and a Gilson Model 112 UV monitor operating at 280
nm using an analytical TSK 3000 column eluting with PBS,
pH=7.2 at 1 mL/min.
28. Yang, D.; Kuan, C. T.; Payne, J.; Kihara, A.; Murray, A.;
Wang, L. M.; Alimandi, M.; Pierce, J. H.; Pastan, I. Clin.
Cancer Res. 1998, 4, 993.
29. Pardrige, W. M.; Buciak, J.; Yang, J.; Wu, D. F. J. Pharm.
Exp. Therap. 1998, 286, 548.
24. Ho, D. K., personal communication.
25. Typical synthesis of 17-GMB-ABA-GA: GA (0.5 g, 0.89
mmol) was dissolved in CHCl3 (300 mL and 1,4-diaminobutane
30. The error bars for the data in Figure 4 do not appear due
their small size. The standard error for all of the data points
therein did not exceed 5%.