Geldanamycin (GM) is a natural antibiotic that binds Hsp90 and induces the degradation of receptor tyrosine kinases, steroid receptors, and Raf. It is a potent inhibitor of cancer cells that overexpress HER-kinases, but its effects on other important proteins may cause significant toxicity and limit its clinical use. We report the synthesis and identification of a GM dimer, GMD-4c, which had selective activity against HER-kinases. Selectivity was a function of linker length and required two intact GM moieties. GMD-4c is a potent inducer of G1 block and apoptosis of breast cancer cell lines that overexpress HER2, but does not appreciably inhibit the growth of 32D cells that lack HER-kinases. GMD-4c could be useful in the treatment of carcinomas dependent on HER-kinases.
HER-family transmembrane receptor tyrosine kinases play an important role in transducing extracellular growth signals and when activated can be oncogenic (1, 2). Overexpression of HER1 and HER2 occurs in a variety of human malignancies. In breast cancer,overexpression of HER2 is associated with a poor prognosis(2). HER1 and HER2 are attractive targets for therapeutic development. Antibodies against each of these receptors have been shown to have antitumor effects in animal models (3). Recently,an anti-HER2 antibody was shown to be effective in the treatment of breast cancers in which HER2 is overexpressed (2, 4). However, therapeutic effects were seen in only a minority of patients and were usually short-lived. Other, more effective methods for HER2 inhibition are needed.
GM3and herbimycin A are benzoquinoid ansamycin antibiotics(5). This class of drugs binds to a specific pocket in the chaperone protein Hsp90 (6, 7). Occupancy of this pocket by the drug leads to the degradation in the proteasome of a subset of proteins that require Hsp90 for conformational maturation(8, 9, 10, 11). These include the HER- and insulin-receptor families of tyrosine kinases, Raf-1 serine kinase, and steroid receptors. The addition of GM to tumor cells leads to a Rb-dependent G1 growth arrest and apoptosis.4HER-kinases are the most sensitive targets of GM, and tumor cell lines in which HER2 is overexpressed are inhibited by especially low concentrations of the drug (12, 13, 14). These findings imply that GM and related drugs may be useful in the treatment of a variety of tumors. An analogue of GM, 17-allylaminoGM, is currently under Phase I clinical trials. However, the number of important signaling molecules that are affected by ansamycins suggests that they may have untoward toxicity.
We have endeavored to synthesize derivatives of GM that have a narrower spectrum of action and greater selectivity. Signaling via the HER-kinases may require their association with Hsp90. This chaperone is required for sevenless (Drosophila epidermal growth factor receptor-family member) signaling (15). v-Src associates with Hsp90 and is very sensitive to GM(6). Both the sensitivity of v-Src to ansamycins and its association with Hsp90 depend on the presence of the catalytic domain but do not require catalytic activity (16, 17). The sensitivity of HER2 to GM also requires the catalytic domain(12). However, a direct interaction of Hsp90 and HER-kinases has not been convincingly demonstrated.
These data suggested to us that Hsp90 is likely to interact with the catalytic domain of HER-kinases. Because HER-kinases undergo dimerization on activation, we speculated that each element of the HER-kinase dimer interacts with Hsp90. Accordingly, it seemed possible that a GMD might be able to interact with both subunits of the HER-kinase dimers. Here, we report the synthesis and evaluation of several GMDs and the identification of a GMD, GMD-4c, which induces the selective degradation of HER-kinases.
Materials and Methods
The human breast cancer cell lines MCF-7 and SKBR-3 were obtained from American Type Culture Collection (Manassas, VA) and maintained in DME/F12 (1:1) supplemented with 10% heat-inactivated FBS (Gemini Bioproducts), 2 mm glutamine, and 50 units/ml each of penicillin and streptomycin, in a humidified 5%CO2/air atmosphere at 37°C. The murine hematopoietic cell line 32D was kindly provided by Dr. Yosef Yarden(The Weizmann Institute of Science, Rehovot, Israel) and maintained in RPMI 1640 supplemented with 10% heat-inactivated FBS, 2 ng/ml interleukin 3 (R&D Systems, Inc.), 2 mm glutamine,and 50 units/ml each of penicillin and streptomycin.
Polyclonal antibodies against HER2 (c-18), HER3 (C-17), Raf-1 (c-12),and PI3 kinase (p85) (Z-8) were purchased from Santa Cruz Biotechnology, Inc. The HER2 monoclonal antibody (Ab-5) for immunoprecipitation was from Oncogene. A monoclonal antibody against ER(clone H-151) was from StressGen Biotechnology Corp. A polyclonal antibody against the α-subunit of IGF-IR was kindly provided by Dr. L-H. Wang (Mt. Sinai Medical Center,New York, NY).
GM and Its Analogues.
GM was kindly provided by Drs. David Newman and Edward Sausville (Drug Synthesis and Chemistry Branch, National Cancer Institute, Bethesda,MD), dissolved in 100% DMSO, and stored at −20°C. The GM analogues were prepared according to S. Kuduk et al.5using the method of Schnur et al. (13). Briefly, the GMDs were prepared by treatment of GM with 0.5 equivalent of the appropriate diamine in DMSO. The ansa-ring-opened GMDs (GMD-a and GMD-aa) were prepared by methanolysis(NaOMe/methanol) of the GMD-4c. GM-quinone was synthesized by first treating GM with excess 1,4-diamobutane, then the addition of 2-methoxy-1-hydroxymethylquinone.
Cell Growth Experiments.
Cells were plated in 6-well tissue-culture plates (Corning Glass) at 20,000 cells/well. Two days after plating, cells were treated with different concentrations of drugs or the vehicle DMSO (0.1%). MCF-7 and SKBR-3 cells were treated for 4 days. Medium with the appropriate drug or vehicle was changed every 2 days. Cells were trypsinized,collected, and counted on a Coulter counter. IC50for cell growth is designated as the amount of each drug needed to inhibit cell growth by 50% compared with the control vehicle. 32D cells were treated and counted every day for 3 consecutive days using a hematocytometer.
Immunoprecipitation and Immunoblotting.
MCF-7 cells were washed twice with ice-cold PBS, collected by scraping,and transferred into microcentrifuge tubes. For immunoblotting (HER2,HER3, Raf-1, ER), cells were lysed with SDS lysis buffer [50 mm Tris-HCl (pH 7.5), 2% SDS, 10% glycerol, and 1 mm DTT], boiled for 10 min, and sonicated briefly. For immunoprecipitation (HER2 and IGF-IR), cells were lysed with NP40 lysis buffer [50 mm Tris-HCl (pH 7.5), 1% NP40, 150 mm NaCl, 1 mmNa3VO4, 40 mmNaF, 1 mm phenylmethylsulfonyl fluoride, and 10 μg/ml each of aprotinin, leupeptin, and soybean trypsin inhibitor] for 20 min at 4°C. Cell lysates were cleared by centrifugation at 14,000 × g for 15 min at 4°C in a microcentrifuge. Supernatants were collected as the experimental samples. Protein concentration in each sample was determined using the BCA kit (Pierce Chemical Co.), according to the manufacturer’s instructions. For detecting IGF-IR, samples were immunoprecipitated with anti-IGF-IR antibody. Immunocomplexes were collected on protein A-Sepharose beads (Pharmacia) and washed three times with the lysis buffer. Samples were subjected to SDS-PAGE, electrotransferred to nitrocellulose membranes, detected using the ECL kit (Amersham Corp.),according to the manufacturer’s protocol, and quantitated using the Gel Doc 1000 (Bio-Rad Laboratories). IC50 for protein degradation is designated as the amount of each drug needed to decrease 50% of the protein (HER2 or Raf-1) compared with the control in MCF-7 cells after a 24-h treatment.
Pulse-labeling and Pulse-chase Experiments.
To study the effects of GM and GMD-4c on protein synthesis, MCF-7 cells were pulse-labeled with [35S]protein-labeling mix (NEN; 100 μCi/ml, 1175 Ci/mmol) in methionine/cysteine-free media for increasing amounts of time in the presence of either GM (1μ m), GMD-4c (1 μm), or the carrier (DMSO,0.1%). To study the effects of GM and GMD-4c on protein degradation,MCF-7 cells were pulse-labeled to isotopic equilibrium with[35S]protein-labeling mix (NEN; 100 μCi/ml,1175 Ci/mmol) in methionine/cysteine-free medium mixed with regular medium/5% FBS (9:1) for 14 h and chased with unlabeled methionine/cysteine (150 μg/ml) in the presence of either GM (1μ m), GMD-4c (1 μm), or the carrier (DMSO,0.1%) for a period of 12 h. Cells were collected at different time points from the pulse-labeling and pulse-chase experiments and lysed with NP40 lysis buffer. Samples containing equal amounts of protein (300 μg) were immunoprecipitated for HER2 and subjected to 7% SDS-PAGE. Gels were dried and exposed to X-ray films. HER2 bands were quantitated using Bio-Rad Gel Doc 1000.
Cells were grown and treated on fibronectin-coated coverslips placed in multiwell plates. Cells were fixed for 20 min at −20°C in 100%methanol, rehydrated in PBS for 10 min at room temperature, and blocked for 30 min at 37°C in a blocking solution consisting of 2% BSA, 10%normal goat serum, and 0.05% Tween 20 in PBS. HER2 and Raf-1 were immunodetected with antibodies (SC284 and SC133, respectively) from Santa Cruz Biotechnology, Inc., and IGF-IR was immunodetected with an antibody (Ab1) from Calbiochem Oncogene Science. Slides were incubated with a 1:1000 dilution of the primary antibody in blocking buffer for 1 h at room temperature, followed by an incubation with Alexa 546-coupled goat antirabbit IgG secondary antibody (A-1 I010; Molecular Probes) at a 1:50 dilution in blocking buffer, or fluoresceine isothiocyanate-coupled goat antimouse IgG (F276; Molecular Probes) at a 1:100 dilution. Slides were washed three times with 1 ml of 0.5% BSA in 0.05% Tween 20 in PBS in between and after incubations with primary and secondary antibodies. DNA was stained with bisbenzimide, which was included in the secondary antibody solution (3 μg/ml final concentration). Coverslips were mounted on glass slides using Vectashield (Vector Laboratories, Inc.) to prevent quenching of fluorescence. Immunofluorescence was detected with a Zeiss epifluorescence microscope at ×40 and ×100, using appropriate filters for detection of rhodamine and bisbenzimide, and via confocal microscopy.
Results and Discussion
We synthesized a series of GMDs covalently joined by alkylamino linkers of varying chain length (Fig. 1). The linker is bonded to the 17-carbon of both GM moieties. The crystal structure of GM bound to Hsp90 shows that the 17-carbon is the only one not buried in the binding pocket (7). The activities of GMDs were compared with those of GM and assessed in terms of efficiency of induction of down-regulation of the HER2 and Raf-1 protein kinases in MCF-7 cells (Table 1). The properties of these GMDs were found to vary as a function of chain length. GM itself causes the induction of HER2 degradation with an IC50 of 45 nm. Dimers with linkers of four to seven carbons retain activity against HER2(IC50, 60–70 nm). Dimers with longer linkers lose activity; the 12-carbon-linked compound has an IC50 of 750 nm.
The four-carbon-linked dimer, GMD-4c, has selective activity. GM causes Raf-1 degradation with an IC50 of 200 nm. GMD-4c is much less active, with an IC50 of 2200 nm. Selectivity is lost with increasing chain length; the seven-carbon-linked-dimer (GMD-7c)retains activity against HER2 (IC50, 70 nm) and is only slightly less active than GM against Raf-1(IC50, 500 nm). As linker carbons increase to more than eight, activity against both targets declines in parallel. The properties of GMD-4c were examined in greater detail. GM causes the degradation over time of HER-kinases, Raf-1, the ER, and,more slowly, the IGF-IR (Fig. 2 %Thus, GMD-4c induces the selective degradation of HER-family kinases and specifically inhibits the growth of HER-kinase containing tumor cell lines. Because its effects on other key signaling proteins are attenuated, GMD-4c is likely to be much less toxic than GM. This work supports the idea that selective ansamycins with a different, more restricted spectrum of targets than the parent molecules can be synthesized. In this case, the mechanism of selectivity is not yet known, but depends on the presence of both GM moieties and is a function of the linker length. GMD-4c may selectively interact with HER-kinase dimers, but it is also possible that it preferentially interacts with different Hsp90-family members than GM. This work represents a new strategy for abrogating growth receptor function in human tumors.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Supported by United States Army Breast Cancer Research Program Grant P01CA68425 (to N. R.), by National Cancer Institute Breast Specialized Program of Research Excellence Grant P50CA68425-02 (to N. R. and L. S-L.), and by NIH Grant CA-28824 (to S. J. D.). F. F. Z. is the recipient of the Ethol Abbot Fellowship.
The abbreviations used are: GM, geldanamycin;GMD, GM dimer; ER, estrogen receptor; IGF-I, insulin-like growth factor I; IGF-IR, IGF-I receptor; FBS, fetal bovine serum; RB, retinoblastoma protein.
M. Srethapakdi and N. Rosen, manuscript in preparation.
S. D. Kuduk et al., unpublished data.
|Druga .||MCF-7 .||.||.||SKBR-3 IC50c (growth inh.) (nm) .|
|.||IC50b (HER-2) (nm) .||IC50b (Raf-1) (nm) .||IC50c (growth inh.) (nm) .||.|
|Druga .||MCF-7 .||.||.||SKBR-3 IC50c (growth inh.) (nm) .|
|.||IC50b (HER-2) (nm) .||IC50b (Raf-1) (nm) .||IC50c (growth inh.) (nm) .||.|
See Fig. 1 for the structure of each drug.
MCF-7 cells were treated with various concentrations of each drug; IC50 for each protein is designated as the amount of each drug needed to decrease the steady-state level of either HER2 or Raf-1 to 50% of the control after 24 h of treatment.
MCF-7 and SKBR-3 cells were treated with different concentrations of each drug for 4 days; IC50 is designated as the amount of each drug needed to inhibit cell growth by 50% compared with the control after the 4-day treatment. The numbers were the average of three different experiments.
We are indebted to Drs. David Newman and Edward Sausville(Developmental Therapeutics Program, National Cancer Institute,Bethesda, MD). We appreciate the assistance of Susan Krueger (Center for Biomedical Imaging Technology, University of Connecticut Health Center, Storrs, CT) with confocal microscopy. Valuable discussions were provided by members of the Rosen’s and Danishefsky’s laboratories, in particular by Dr. Ouathek Ouerfelli.