Targeting the folate receptor (FR): imaging and cytotoxicity of ReI conjugates in FR-overexpressing cancer cells.

DOI:10.1002/cmdc.200800125 Targeting the Folate Receptor (FR): Imaging and Cytotoxicity of ReI Conjugates in FR-Overexpressing Cancer Cells Nerissa Viola-Villegas, Amy E. Rabideau, Justin Cesnavicious, Jon Zubieta, and Robert P. Doyle*[a] The synthesis, characterization, invitro imaging, and cytotoxic ovary cell line in which no internalization of g-4 was observed. properties of a new folate conjugate of rhenium(I) are reported. g-4was found to be cytotoxic with IC values of 189 and 78mm 50 The conjugate [FA-PEG-BQAV-Re(CO)]+ ACHTUNGTRENNUNG(g-4) was screened at6and24h,respectively,towardtheFR-positivecellline.Thisis 3 against an adriamycin- and cisplatin-resistant human ovarian in contrast to the IC value of 502mm at 6h and 84mm at 24h 50 cancer cell line (A2780/AD) that overexpresses the folate receptor forcisplatininthesamecellline,withasignificantlygreater tox- (FR). Compound g-4 was internalized by a folate-receptor-medi- icityattheearliertimepoint.Thecytotoxicityofg-4asexplained atedendocytoticpathway,whichresultsininternalaccumulation by interactions that occur between the rhenium(I) complex of g-4. This was contrasted with a FR-negative Chinese hamster moietyandDNAisdescribed. Introduction Specific targeting of cancer cells remains one of the major antiproliferative compounds or imaging agents to locate such themes in medicinal chemistry. The goal is one of overcoming cancerous tissue invivo, and a number of examples of such a the problems that are associated with the loss of healthy usehavebeenreported. tissue by using classic chemotherapy and has driven the Low etal. have investigated a 67Ga–deferoxamine complex search for, and identification of, unique markers of tumors that is conjugated to FA,[5] while other groups have looked at such as prostate-specific membrane antigen or overexpressed FA-based conjugates of metal radionuclide agents such as: folate receptor (FR). The ability to selectively target tumors 66/68Ga,[6] 111In,[7] and 64Cu.[8] The FR has also been exploited for would be predicted to assuage patients’ treatment burden in single-photon emission computed tomography (SPECT) by at- thatitprovidesatropismforthedrugandgreatlyimprovesits taching a 99mTc–diethylenetriamine pentaacetic acid complex therapeutic index. The useof overexpressed FR in particular as withapolyethyleneglycolspacertoFA.[9]Thisresultedinlocal- a tumor marker has generated considerable interest.[1] Target- ized imaging in lymphatic tumors that express FR, and low ing the FR is attractive because in addition to being overex- uptake in other tissues except for the kidneys.[9] Recently, a pressed in tumor lines, it is down-regulated (and inaccessible 99mTc/ReI tricarbonyl core that is chelated by a picolylamine to blood circulation) in healthy adult cells. The FR facilitates monoacetic acid with an aminohexane spacer and conjugated theuptakeoffolicacid(FA)(seeFigure1),avitamin(B)thatis to FA has been reported.[10a] In this investigation, Schibli and 9 necessaryforcellgrowthandproliferation.[1]TheFRhasahigh co-workersshowedeffectivecellbinding andinternalization of binding affinity (K ~0.426109m) for FA,[2] and once bound, FA thefolatederivativesofthemetalfluorophore.[10a] D is rapidly delivered into the cell through endocytosis.[3] Several This widespread use of 99mTc in diagnosticmedicinehas also tumors overexpress the FR, including ovarian (A2780), breast drawn attention to the potential of technetium’s heavier con- (MDA-MB-231), cervical (HeLa-IU), nasopharyngeal (KB), and gener, rhenium and in particular isotopes 186Re and 188Re. For 1 colon (Caco-2).[4] Consequently the FR is ideal for delivery of example, ReI has been complexed to boronic acid derivatives ofbipyridineasaninvivoopticalglucosedetectortoeliminate the need for blood sampling to determine blood glucose lev- els.[10b] This approach is based on fluorescence switching through photoinduced electron transfer, in which the absence of glucose quenches the fluorescence of the metal complex [a] N.Viola-Villegas,A.E.Rabideau,J.Cesnavicious,J.Zubieta,R.P.Doyle DepartmentofChemistry,SyracuseUniversity Syracuse,NY13244-4100(USA) Fax:(+1)315-443-4070 E-mail:rpdoyle@syr.edu Figure1.FAwithitsthreemajorstructuralcomponentsincludingthea- SupportinginformationforthisarticleisavailableontheWWWunder andg-carboxylicacidgroupoftheglutamatemoietyindicated. http://dx.doi.org/10.1002/cmdc.200800125. ChemMedChem2008,3,1387–1394 H2008Wiley-VCHVerlagGmbH&Co.KGaA,Weinheim 1387 MED R.P.Doyleetal. transducer and the presence of glucose increases lumines- cence intensity.[10b] 188Re (t =17h, e =2.12MeV, e =155keV) 1/2 b g has also been explored mainly for its radiotherapeutic proper- ties. Arecent study targeted metastatic melanomas by using a heptapeptide conjugate of the radionuclide, 188Re-HYNIC-Asn- Pro-Asn-Trp-Gly-Pro-Arg (HYNIC=6-hydrazinonicotinamide), which resulted in considerable tumor retardation.[11] Recently, 186/188Re systems that were complexed to 1-hydroxyethylidene- 1,1-diphosphonate showed promise as a radiopharmaceutical for the palliative treatment of bone cancers, with patients re- porting significant pain relief.[12] Furthermore, Spitzweg and colleagues have explored the therapeutic properties of 188Re, whichistransportedbyasodiumiodidesymporter andresults intumorvolumedecreaseinprostratecancer.[13a]Thefirst188Re radiofolate was prepared and its pharmacokinetic profile was investigated alongside with 99mTc.[13b] Biodistribution experi- ments of the 188Re conjugate showed excellent tumor uptake with low tumor-to-kidney ratios upon administration of an an- tifolate.[13b] Combining then the intriguing imaging and therapeutic properties of rhenium isotopes with the molecular target that Scheme1.Synthesisofg-1viaDCC/NHScoupling:a)DCC/NHS,DMSO is offered by the folate receptor could produce effective phar- (anhyd);b)HN-PEG-NH,pyridine,DMSO(anhyd). 2 2 maceuticalsforthediagnosisand/ortreatmentofFR- positive cancers. Based on these considerations, we havesetouttoinvestigatethepropertiesofabisqui- noline-based ReI chelate complex bound to FA–poly- ethylene glycol conjugates. Herein we present the synthesis and characterization of a new ReI conju- gate, and demonstrate its selectivity for FR cell lines anditsintriguingcytotoxicproperties. Results and Discussion Synthesisandcharacterization Scheme2.Preparationof2viareductiveamination:a)Na[BHACHTUNGTRENNUNG(CHCOO)]. 2 3 The synthesis of g-1 was performed by coupling of polyethylene glycol (PEG, M~2000) to the glutamate moiety trospray ionization mass spectrometry (ESI-MS, see Supporting r ofFA.FAwasactivatedbyN,N’-dicyclohexylcarbodiimide(DCC) Information), and the 1HNMR spectrum was consistent with 2. and N-hydroxysuccinimide (NHS) coupling agents as shown in The complexation of [Re(CO)ACHTUNGTRENNUNG(HO)]+ to 2 (see Scheme3) 3 2 3 Scheme1. The reaction yielded two FA-based regioisomers, was carried out in methanol by reflux at 708C for 16h. Com- designated a- and g-1. These isomers were separated by ion- pound 3 was isolated by using reversed-phase C HPLC (RP- 18 exchangechromatographyandcharacterizedasdescribedpre- HPLC); it had a retention time of t ~10.2min, in contrast to R viously.[14] Such separation is necessary to remove the inactive that of 2, which eluted at ~6.0min. Compound 3 was charac- a-1system,whichisnotrecognizedbytheFR.[15] terized by 1HNMR spectroscopy and ESI-MS. Furthermore, the Compound 2 was prepared by reductive amination of two isotopic distribution pattern observed in ESI-MS was also con- moles of quinoline-2-carboxalde- hyde and the amino acid linker, 5-aminovaleric acid as previously described and shown in Scheme2.[16] A modification in the literature purification step was made by precipitating 2 by using ethyl acetate/hexane (2:8) at room temperature. This elimi- natedtheneedforcolumnchro- matography. The expected mass (m/z 399) was verified by elec- Scheme3.Labelingof2with[Re(CO)]+:a)MeOH,708C. 3 1388 www.chemmedchem.org H2008Wiley-VCHVerlagGmbH&Co.KGaA,Weinheim ChemMedChem2008,3,1387–1394 ReIConjugatesforCancer sistent with rhenium incorporation (see Supporting Informa- within ~3min.[17] Uptake of g-4 was visualized by fluorescence tion). Electronic absorption spectroscopy provided molar ex- in A2780/AD cells (see Figure2). Scanning from the surface tinction coefficients (e) at 232nm (108625m(cid:1)1cm(cid:1)1) and through the intracellular regions to determine internalization 314nm (23776m(cid:1)1cm(cid:1)1), which are assigned to p!p* transi- of the drug was performed by slicing through the cell in 1mm tionsfromthearomaticquinolinerings. increments (see Supporting Information). Fluorescence was Compound g-1 was conjugated to 3 via a 1,1’-carbonyl-di- uniformly distributed in and throughout the cell, which clearly (1,2,4-triazole) (CDT) coupling reaction in dry DMSO as shown indicatesthesuccessfuluptakeofg-4.Nointernalizationofg-4 in Scheme4; 3 was activated by CDTat 608C over 30min. A inCHOwasobserved,althoughsomenonspecificsurfacebind- ingwasnoted.Thisisduetothe presence of PEG and has been reported.[17] Performing an acidi- fied saline wash decreases this occurrence.[18]Scanningfromthe surface through the cell in 1mm increments was also performed ontheCHOcells,andnofluores- cence was observed. This is con- sistent with no internalization of g-4 in CHO cells; g-4 is bound only to the cell surface mem- brane(datanotshown). To determine whether the presence of FA inhibited the binding of g-4 to the FR, an excess of FA (~500mm) was added to a solution of g-4 (10mm) and incubated, for be- tween 45min and 24h, with the Scheme4.Conjugationofg-1and3viaCDTcoupling:a)CDT,608C,30min;b)TEA. FR overexpressing A2780/AD cells growing in FA-containing RPMI1640 media. No uptake yellow solution of g-1 was then added to this solution. An was observed as followed by confocal microscopy in contrast amber-colored solution was observed after 16h. 1,2-Dichloro- to the rapid and extensive internalization noted in A2780/AD ethane was added to this solution, and the solution was cells that were grown in the absence of FA. This is consistent washed with water. The organic layer was dried invacuo and with the idea that binding of g-4 to the FR is completely redissolved in water/acetonitrile (1:9). Purification was carried blocked by the presence of free FA. Even after 24 hours of in- outwithRP-HPLC.Compoundg-4elutedataretentiontimeof cubation with g-4 and excess FA, no internalization was ob- t ~13.4min. served. R The identity of g-4 was confirmed by MALDI-TOF MS; the compound displayed a mass peak that was centered at m/z Invitrocellcytotoxicity ~3007, which corresponds tothe desired theoreticalmass (see Supporting Information). The broad variance is the result of To the best of our knowledge, there has been limited investi- PEG polydispersity (peak separation of m/z 88 corresponds to gationoftheantiproliferativepropertiesofReIconjugates,and twoethyleneglycol units(m/z44)).Further characterizationby only limited reports of IC values for ReI compounds.[19,20] We 50 1HNMR spectroscopy was also consistent with the identity of have conducted antiproliferative studies to determine the g-4. impact of the presence ReI in terms of cytotoxicity. Com- pounds g-1, 2, 3, and g-4 were tested against the FR-overex- pressingA2780/ADandFR-freeCHOcelllinesforsuitablecom- Invitrocellimaging parison over a time period of 6 and 24h (see Table1). Results Confocal microscopy experiments were performed on FR-over- indicatethatboththerhenium-freesystems,namelyg-1and2, expressing A2780/AD cells, with CHO cells as negative control have low toxicity, marked by the high-micromolar IC values 50 (CHO cells do not express FR, as confirmed by RT-PCR). Com- that were determined for 2 and by the fact that no IC value 50 pound g-4wasadded tothecells ataconcentration of10mm, was notedforg-1uptoaconcentrationof5mm(seeTable1). and the solution was incubated for 45min and 24h. No time The cytotoxicity on either of the two lines, FR-free or FR-over- dependencewasnotedinthisrange,andanincubationperiod expressingcellswasalsosimilarforbothg-1and2.Incontrast, of 45min was sufficient for uptake and internalization; this is 3andg-4wereshowntohavesignificantlygreatertoxicity,es- consistent with FR’s known ability to bind and internalize FA pecially in the earlier 6-hour time point, for FR-expressing cells ChemMedChem2008,3,1387–1394 H2008Wiley-VCHVerlagGmbH&Co.KGaA,Weinheim www.chemmedchem.org 1389 MED R.P.Doyleetal. than FR-free cells (see Table1). Clearly the presence of the FR and the presence of rhenium itself are playing a key role in toxicity. The greater toxicity of g-4 over 3 might be related to uptake and cellular retention. Compound g-4 undergoes facili- tated transport through the FR, but 3 can only gain entry through passive diffusion. This facilitated uptake and the fact that FA–PEG-containing systems have been shown to have in- creased intracellular residency times[21] would explain this greater antiproliferative activity. This postulated increased resi- dency might also play a role when comparison is made be- tweenthetoxicityofg-4andcisplatininA2780/ADcells;g-4is significantly more toxic than cisplatin in the multi-drug-resist- ant A2780/AD cell line at 6h. This can be rationalized by the efficient uptake of g-4 via the FR endocytotic pathway, which evidently out-competes the rate of intracellular cisplatin accu- mulation. IC values werealsodeterminedfornon-FR-expressingCHO 50 cells (see Table1). As in the A2780/AD cells, the presence of rhenium has significantly increased the cytotoxicity over g-1 and 2. Also of interest here is the observed toxicity, which is similar to that observed in the FR-expressing A2780/AD cells, for g-4 (189 and 204mm at 6h and 78.2 and 80.8mm for A2780/ADandCHOrespectively)inCHOcells.Thetoxicitythat is associated with 3 can be explained by passive diffusion for cellularentry,but,asevidencedbyconfocalmicroscopyexperi- ments, nonspecific cell surface binding was observed for g-4, and no internalization was observed after 45min and 24h in- cubation. Toxicity then might be due to these surface interac- tions. Previous work describing the effects of metal ions on membranesurfaceshascertainlyshowntoxicitycanoccur,[22–24] but the exact effects that are induced by g-4 on the mem- brane surface requires further investigation. Finally, unlike the resistance to cisplatin that is observed in A2780/AD cells, cis- platin had IC values in the low mM range (12 and 4.8mm at 6 50 and 24h respectively, compared with 502 and 84mm for the sametimepointsinA2780/ADcells). Ligandchallengeexperiments Competitive binding experiments were performed with a 100- fold excess of histidine (100mm) or 5mm glutathione on 1mm 3 in phosphate-buffered saline (PBS) at pH7.4 to gauge Figure2.A2780/ADcellsincubatedwithg-4:a)106 magnificationshowing the stability of 3. This is important given that the presence of fluorescentlylabeledcells;b)636 magnificationshowingacloserlookof rhenium clearly impacts the antiproliferative activity of the thecellularpool. compounds and noting that rhenium release would aid to our understanding of mechanism. Three solutions were used: 1)3 in PBS, 2)3 in PBS with histidine and 3)3 in PBS and gluta- thione. The solutions were incubated at 378C over a period of Table1. IC valuesinA2780/ADandCHOcells[mm]. 50 48h. Analytical C RP-HPLC runs of all three solutions yielded 18 Drug A2780/AD CHO one major peak with a retention time of t R ~11.8min. No new 6h 24h 6h 24h peaks were observed over 48h aside from a new peak in the g-1[a] – – – – glutathione run in which a peak at 3.77min was observed 2 2837(cid:3)0.310 1765(cid:3)0.0400 2979(cid:3)0.210 1526(cid:3)0.0200 (datanotshown).ThispeakwascollectedandanalyzedbyESI- 3 1684(cid:3)1.80 590(cid:3)5.88 2627(cid:3)2.80 746(cid:3)3.90 MS and inductively coupled plasma analysis (ICP). Mass spec- g-4 189(cid:3)15.0 78.2(cid:3)0.210 204(cid:3)12.0 80.8(cid:3)13.0 trometry indicated this new peak was simply oxidized gluta- cisplatin 502 84 12 4.8 thione and ICP confirmed that no rhenium was present in the [a]Notcytotoxicatconcentrationsupto2mm. sample (data not shown). These observations therefore sup- 1390 www.chemmedchem.org H2008Wiley-VCHVerlagGmbH&Co.KGaA,Weinheim ChemMedChem2008,3,1387–1394 ReIConjugatesforCancer portthestabilityofthebisquinolineligandinbindingrhenium undertheseconditions. DNAgelmobilityshiftassays pUC19DNA (15nm) was incubated with 3, over 24 to 48h, withmetal-ligand-to-DNAratios(r )rangingfrom1.0to5.0.No b scission of the DNA was observed as shown in Figure3. No effect on DNA mobility was observed either based on control, with no change in the open circular (oc), or supercoiled (sc) DNA. Figure4.UV/Visabsorptionspectrumof3withpUC19DNAat378Cover 24hshowingadecreaseinabsorptionindicativeofDNAintercalation. ces the interaction with the negatively charged helix. Work by Figure3.DNAmobilityshiftassaysofincubationtimepointsofa)24and b)48h,showingthetwoformsofDNA(ocandsc)presentinthecontrol Schneider etal. has shown that destacking of nucleobases by andwiththepresenceof3.Lane1:[f+l]DNAmarker(72bp–23kbp small aromatic systems(e.g., quinoline) is promoted by aposi- marker);Lane2:DNAcontrol;Lane3:r b =1.0;Lane4:r b =2.0;Lane5: tive charge in the ligand system.[30] In addition to intercalation r =5.0.Forthegelat48h,DNAbands1,2,3,and4werecollected,purified b is the possibility of minor-groove binding. Minor-groove bind- andanalyzedforrhenium(I)ICP.AllbandsweredeterminedbyICPtohave nodetectablelimitsofrhenium. ers have structural motifs that are i)crescent-shaped ii)posi- tively charged and iii)moderately rotational to fit in the minor groove of the nucleic acid.[31] The crescent form corresponds Interestingly, Alberto etal. reported that an equilibrium well with the helicity of the minor groove.[32] These previous exists between the ocand scforms ofpDNA after exposure to reports can be correlated to the mechanistic activity of 3. A ReI prodrugs.[25] Rhenium was reported to bind to N7 in gua- closer look at the solid state structure of 3 shows a crescent- nine with high kinetic stability as elucidated by its solid state shaped structure with the bond angle of 77.728 between the structure.[26] It was still not conclusive then whether 3 was in- rhenium(I) atom and the nitrogen atoms on the quinoline teracting with DNA. To bridge this gap, intercalation studies rings.[33] The structure of 3 is then consistent with the criteria wereconductedbyusingelectronicabsorptionspectroscopy. forminor-groovebinding toDNAwiththequinolineringspre- ferring the A+Tsites of the helix, and with the positive charge contributed by the metal ion. The toxicity of 3 might then be DNA Intercalation Study attributed to interactions between DNA and 3. To investigate Electronic absorption spectroscopy of 3 in the presence of whether either this interaction or intercalation led to topoiso- 0.3mg pUC19DNA at 378C over 24h was performed (see merase inhibition, we also performed a topoisomerase inhibi- Figure4). A slight hypochromicity that was associated with a tionassays. 0.021 decrease in absorption can be attributed to intercalation interactions of 3 with the DNA base pairs. Studies showing Topoisomerase I Inhibition Assay alkyl-linked diquinolines have reported selective binding to A+T-rich regions of the DNA.[27] The binding affinity was Many anticancer therapeuticsinterferewiththeactivity oftop- shown to increase as the alkyl chain length increased.[27] Luzo- oisomeraseIandII(TopIandII).Topoisomerasesareisomerase peptin, which contains two quinoline moieties that are linked enzymes that act on the topology of DNA, and are critical for by a cyclic depsipeptide was shown to bifunctionally interca- theunwindingofDNAthatisnecessaryforreplication.Cancers late and induce DNA cross-linking.[28] In addition, Crooke etal. of the kidney,[34] colon,[34] prostrate,[34] ovary,[35] and esopha- reported that the half molecule of luzopeptin (i.e., quinoline gus[36] have been determined to have elevated topoisomerase with a pentapeptide chain) can effectively interact with DNA levels, which is a necessary consequence of rapid and un- butdoesnotintercalate.[28] checked proliferation. TopI assays with 3 on plasmid fX174, Inanotherstudyofquinoline-derivatizedechinomycin(i.e.,a which contains both oc and sc DNA indicated that the ability naturally occurring quinoxaline antibiotic), 2QN was reported of TopI to unwind sc into oc DNA was not inhibited by con- to behave as a bisintercalator with preference to polyACHTUNGTRENNUNG(dA–dT) centrationsof3ofupto5mm(seeFigure5).Clearly,noinhibi- rich sites.[29] In addition, 3 is positively charged, which enhan- tion of the TopI activity was observed. No re-supercoiling of ChemMedChem2008,3,1387–1394 H2008Wiley-VCHVerlagGmbH&Co.KGaA,Weinheim www.chemmedchem.org 1391 MED R.P.Doyleetal. fluoroacetic acid (TFA, 99%, Aldrich). [Re(CO)ACHTUNGTRENNUNG(HO)]+ was pre- 3 2 3 paredbyfollowingamethodthatwasdescribedpreviously.[38] Forinvitrocellstudies,allmanipulationswereperformedinaLab- conco PurifierI laminar flow hood that had been disinfected with 70%ethanolandirradiatedwithUVlight.CHOcellswereobtained from the American Type Culture Collection (ATCC). The A2780/AD cell line that was used for testing was provided by the Fox Chase Cancer Centre, Philadelphia (USA) and was generated in 1984. Fetal bovine serum (FBS) was purchased from Hyclone. Penicillin– Figure5.TopIinhibitionassaywithfX174RFI.Lane1:lDNAmarker streptomycin solution with 10000 units penicillin and 10mgmL(cid:1)1 (564bp–23.13kbp);Lane2:DNAcontrol;Lane3:DNAwithTopI;Lanes4–8 streptomycin in 0.9% NaCl was obtained from Sigma. Gibco containDNA,TopI,andvariousconcentrationsof3(1,10,100,1000, RPMI1640 16 growth media that contained l-glutamine and 5000mm). phenol red without folic acid was supplied by Invitrogen. Cellgro Cellstripper, a nonenzymatic cell dissociation solution, was ob- tained from Mediatech. F-12K Ham’s media was purchased from theDNAwassubsequentlyobserveduponremovalofthetop- ATCC.Growthmediawerefilteredwith0.45mmfilter(Fisher).2-(2- oisomeraseandtheadditionof3;thisindicatesthat3wasnot Methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H- intercalating, or even weakly intercalating (also suggested by tetrazolium, monosodium salt solution (WSK-8) was purchased Figure4) at concentrations of less than 5mm.[37] A DNA-based from Dojindo. Cells were incubated and grown in an incubator mechanism of cytotoxic action seems unlikely then for this thatwaspurchasedfromVWR.Opticaldensitiesofthecellcultures were measured at 450nm with Thermo MultiskanEX that was system and as with many metal-based drugs, the exact mode equipped with Ascent software version2.6. Confocal microscopy ofactionrequiresfurtherinvestigation. experiments were conducted with a Zeiss LSM4 Pascal confocal microscope and Image Analysis equipped with argon and HeNe laser. Conclusions The following materials were used for DNA experiments: Topoiso- WehavesuccessfullysynthesizedaFA–PEG-derivatizedReIcon- meraseI (TopI) and bovine serum albumin (BSA) were purchased jugate and tested its invitro imaging and cytotoxic properties. from New England Biolabs. Agarose and Tris(boric acid)–EDTA Addition of the FA–PEG moiety to the Re–BQAV complex af- (106) were obtained from Merck. Ethidium bromide and gel-load- fords significant internalization of the complex in FR-express- ingdye(66)werepurchasedfromPromegaandAmresco,respec- tively. fX174 DNA was purchased from Invitrogen. pUC19 was ing cells compared with no internalization in non-FR-express- transformed from chemically competent E.coli and isolated and ing cells; this is consistent with a receptor-mediated endocy- purified by using the Wizard PlusSV Miniprep DNA purification toticuptakemechanism.ThisselectivityofuptakebetweenFR- system from Promega. Gels were viewed by using a Kodak Gel devoid and FR-overexpressing cell lines suggests that g-4 Logic100imagingsystemUVtransilluminator. might be a suitable candidate for diagnostic imaging of FR- overexpressing carcinoma invivo by replacing Re with 99mTc. All other reagents and buffers that were used were of reagent The significant cytotoxicity suggests that it might actually be grade or higher. Ultrapure water (18.2MW) was used through out possible to use this system with 186/188Re as a targeted thera- theinvestigation. peutic in cell lines where drug-resistance is evident. This work An Agilent1100 reversed-phase HPLC instrument with manual in- then could lead to a novel diagnostic radioimaging or thera- jection and automated fraction collector was fitted with a Zorbax peutic agent that targets FR-overexpressing carcinoma invivo C analyticalcolumn(42610mm)foranalyticalruns.Theflowrate 18 andworktoinvestigatethisisunderway. was0.7mLmin(cid:1)1.PurificationwasmadewithaC semipreparative 18 column (9.46250mm) at a flow rate of 2mLmin(cid:1)1. Detection was carried out by UV monitoring at 254nm. The gradient used was Experimental Section 55% 0.1% TFA in H 2 O and 45% 0.1% TFA in MeCN to 40% 0.1% TFAinHOand60%0.1%TFAinMeCNover15min.Ion-exchange 2 Materials:AllreactionswereperformedunderaninertargonorN chromatographywasconductedonanAktaPrimePlusinstrument 2 atmosphere by using Schlenk techniques. Drying of dimethylsulf- byusingPrimeview5.0software.TheANX(1mL)andthePD10Se- oxide (DMSO, 99%, Sigma) was performed by charging through a phadexG-25Mdesalting(10mL)columnswerepurchasedfromGE column of molecular sieves (4P, Mallincrodt) dried overnight at Health Sciences. 1HNMR spectroscopy was performed with Bruker 1208C. The following reagents were purchased and used without AvanceDPX 500MHz and Bruker Avance DPX300 instruments. A furtherpurification:folicacid(98%,Sigma),N,N’-dicyclohexylcarbo- Shimadzu LCMS-2010A mass spectrometer and a Bruker Autoflex diimide(DCC,(cid:4)99%,Fluka),polyoxyethylenebisACHTUNGTRENNUNG(amine)(PEG,M~ matrix-assisted laser desorption ionization time-of-flight (MALDI- r 2000) (Fluka), N-hydroxysuccinimide (NHS, (cid:4)97%, Fluka), 1,1’-car- TOF) mass spectrometer were used for ESI and MALDI-TOF MS bonyldi-(1,2,4-triazole)((cid:4)90%,CDT,Fluka),quinoline-2-carboxalde- analysis, respectively. A PerkinElmer ELAN6100 instrument was hyde (98%, Alfa Aesar), 5-aminovaleric acid (97%, Sigma), sodium usedtoconductinductivelycoupledplasmaanalysis(ICP).AVarian triacetoxyborohydride (95%, Sigma), 1,2-dichloroethane (DCE, Cary50BioUV/Visspectrometerwasusedtorecordtheelectronic (cid:4)99%, Sigma), dichloromethane ((cid:4)99.5%, Sigma), methanol absorption spectra of the compounds and was also used for DNA (MeOH, (cid:4)99.8%, Sigma), acetonitrile (MeCN, (cid:4) 99.8%, Sigma), intercalation studies. Centrifugation was performed for 10min at NaSO ((cid:4)99.9%, anhydrous, Sigma) EtN (99.5%, Sigma), and tri- 4000rpm at 48C with a Sorvall Legend RTcentrifuge. All reactions 2 4 3 1392 www.chemmedchem.org H2008Wiley-VCHVerlagGmbH&Co.KGaA,Weinheim ChemMedChem2008,3,1387–1394 ReIConjugatesforCancer except for 2 and 3 were performed in a darkroom under a 15W line were cultured as adherent monolayers in RPMI1640 growth redlight. media that contained l-glutamine and folic acid and F-12K Ham’s mediarespectively.Thesearesupplementedwith penicillin(10000 SynthesisofFA-PEG-NH 1:Thesynthesisof1wasdescribedpre- 2 units) and 10mgmL(cid:1)1 streptomycin (Sigma), 10% (v/v) FBS viously.[39]Theisomers,a-andg-of1werepurifiedbyusingapro- (Sigma). Prior to testing, cells were then passed in RPMI1640 FA- cedurethatwasestablishedinourresearchgroup.[14] free growth media with FBS and penicillin–streptomycin solution. Synthesis of {[bis(quinolin-2-ylmethyl)amino]-5-valeric acid} It is important to note that cell growth in FA-free media declined (BQAV) (2): Compound 2 was prepared by using a previously es- after several passages. Cells were incubated and grown in a VWR tablishedprocedure.[16]5-aminovalericacid(0.633g,5.4mmol)was mammalian incubator at 5% CO 2 and 95% humidity. Cells were dissolvedinDCE(20mL).Asolutionofquinoline-2-carboxaldehyde culturedinMillipore250mLculturebottleswithventedlids. (1.79g,11.34mmol)inDCE(10mL)wasaddeddropwisetothisso- Confocal microscopy experiments: CHO and A2780/AD ovarian lution. The mixture was stirred for 2h at room temperature and cancer cells (150000 cells/dish) were plated separately on 356 subsequently cooled down to 08C. Sodium triacetoxyborohydride 100mmventeddishes.Thecellswereincubatedat378Covernight (2.63g,12.41mmol)wasslowlyadded,andthemixturewasstirred in FA-free RPMI1640 media. To each plate, 10mm g-4 (1mL) was underN for2h.The solutionwas then extracted with CHCl (36 2 2 2 added and incubated for 45min and 24h. The drug was then re- 50mL) and water (3650mL). The organic layer was washed with 3.4m NaCl (50mL) and dried with NaSO. The organic layer was moved, and the cells were washed with 50mm phosphate buf- 2 4 feredsaline(PBS;361mL).Thecellswerethenwashedwithacidi- thendriedinvacuo.Thedark-redoilysubstancethatwasobtained fied saline solution (361mL; 3.4mm NaCl, pH3.0) and fixed with was redissolved in a solution of EtOAc/hexane (1:1) and the red ice-coldMeOH. precipitate that was formed was then collected by centrifugation at4000rpmover10min;yield:215mg(34.0%).1HNMR(300MHz, For the competitive binding affinity assays with folic acid, A2780/ MeOD, 258C): d=8.28 (d, 3J =8.7Hz, 2H), 7.97 (d, 3J =8.1Hz, H,H H,H ADovariancancercells(150000cells)wereplatedon356100mm 2H), 7.86 (d, 3J =2.4Hz, 2H), 7.77–7.69 (m, 4H), 7.55 (t, 3J H,H H,H vented dishes. The cells were incubated at 378C in FA-containing =7.2Hz, 2H), 4.10 (s, 4H), 2.76 (t, 3J H,H =6.6Hz, 2H), 2.19 (t, 3J H,H RPMI1640media.Themediawasthenremovedanda10mmsolu- =6.6Hz, 2H), 1.66ppm (brm, 4H); UV/Vis (MeCN/H 2 O, 9:1): l max tion of 4ACHTUNGTRENNUNG(1mL) that contained 500mm FA was added. The cells (e)=230nm (95760m(cid:1)1cm(cid:1)1); ESI-MS (10eV, 0.1% TFA in MeCN): were incubated for 45min and 24h. The drug was removed and m/zcalcdforC 25 H 25 N 3 O 2 :399.1;found:400.0[M+H+]. thecellswerewashedwithof50mmPBS(361mL),andthenacidi- Synthesis of [Re(CO)-BQAV]+ 3: [Re(CO)ACHTUNGTRENNUNG(HO)]+ (24.3mg, fied saline solution (361mL; 3.4mm NaCl, pH3.0) to reduce non- 3 3 2 3 0.075mmol) was dissolved in MeOH (20mL). A solution of 2 specific cell surface interactions. The cells were then fixed with (30mg, 0.075mmol) in MeOH (5mL) was added dropwise to the coldMeOH. flask.Thismixturewasheatedtorefluxat708Cunderanargonat- Drug cytotoxicity: The proliferation of the exponential phase cul- mosphereovernight.Thesolutionwasthendriedinvacuoand re- tures of A2780/AD and CHO cells was assessed by WSK-8 colori- dissolvedinCHCl (10mL).HO(3610mL)wasthenaddedtothe 2 2 2 metricassay.WSK-8wasperformedaccordingtomanufacturer’sin- solution, and an extraction was performed. NaSO (~25g) was 2 4 structions (Dojindo). Adherent cell cultures were harvested by added to remove the residual aqueous solvent. The mixture was stripping of culture flasks by using a nonenzymatic cell stripper thendriedinvacuo andredissolved inMeOH.Purification and iso- after a 30min incubation period. The cell densities were then ad- lation of 3 were performed by using C RP-HPLC with t = 18 R justed to 5.06104cellsmL(cid:1)1 for exponential growth over the 10.2min; yield: 12.1mg (40.4%). 1HNMR (300MHz, MeOD, 258C): period of drug exposure. To each well, aliquots (100mL) were ino- d=8.43 (d, 3J =8.4Hz, 2H), 8.02 (d, 3J =8.4Hz, 2H), 7.98 (d, H,H H,H culated,whichresultedin5000cellsperwell.Aftera24-hourincu- 3J =8.4Hz, 2H), 7.82 (t, 3J =8.1Hz, 2H), 7.64 (t, 3J =7.2Hz, H,H H,H H,H bationtimetofacilitateadherence,theFA-freeRPMImediawasre- 2H), 7.53 (d, 3J =8.7Hz, 2H), 4.97–5.25 (m, 6H), 3.67 (d, 3J H,H H,H movedandreplacedwith200mLoffreshmediathatcontaineddif- =1.8Hz, 1H), 3.46 (brs, 5H), 2.48 (t, 3J =7.2Hz, 2H), 2.06 (brs, H,H ferentconcentrationsof2,3,andg-4.Thecellswerethenincubat- 2H), 1.77ppm (m, 2H); UV/Vis (MeCN/HO, 9:1): l (e)=232nm (108625m(cid:1)1cm(cid:1)1), 314nm (3776m(cid:1)1cm 2 (cid:1)1); ESI-M m S ax (10eV, HO/ ed for 6 and 24h. Optical densities were measured by using a 2 plate reader. The percentage of cell viability was determined rela- MeCN, 2:3): m/z calcd for C H NORe: 669.2; found: 669.8 [M+]. 28 25 3 5 tivetountreatedcontrolmicrocultures.TheIC valueswerecalcu- 50 Synthesisof[FA-PEG-Re(CO)-BQAV]+ g-4:Compound3(10.2mg, lated based on an exponential fit by using OriginLabs8 software 3 0.0152mmol) and 1,1’-carbonyl-di-(1,2,4-triazole) (7.48mg, with R2 values (cid:4)0.90 in all cases. All experimental points were 0.456mmol)weredissolvedindryDMSO(1mL)ina50mLSchlenk measuredintriplicateandeachexperimentwasperformedatleast tube.Themixturewasstirredandheatedat608Cfor30min.Aso- three times on separate ’batches’ of compound on cells with no lution of g-1 (24.6mg, 0.0102mmol) in DMSO (1mL) and EtN morethan14passages. 3 (75mL) was added dropwise to 3. The mixture was stirred over- night at room temperature under N. The mixture was then dried Ligand challenge experiments:Three1mmsolutions of g-4 were overnight invacuo. A yellow solid w 2 as collected and was subse- prepared in 50mm PBS buffer (pH7.4). One solution contained quently purified by using C RP-HPLC (t =13.4min); yield: 100mm histidine and another with 5mm glutathione. These solu- 18 R 5.71mg (23.2%). 1HNMR (300MHz, [D 6 ]DMSO): d=8.67 (d, 3J H,H tions were incubated over 48h at 378C. C 18 RP-HPLC analysis was =9.5Hz,2H),8.39(d,3J =9.3Hz,2H),8.11(d,3J =7.8Hz,2H), conductedat0,24and48h.PeakscollectedwereanalyzedbyESI- H,H H,H 7.96 (t, 3J =7.5Hz, 4H), 7.72 (m, 4H), 6.61 (s, 2H), 5.14 (m, 5H), MSandICP. H,H 3.8 (m, 2H), 3.38 (m; PEG), 2.75 (s, 1H), 2.28 (s, 1H), 2.10 (s, 1H), DNAgelmobilityshiftassays:pUC19DNA(15nm)solutionswere 1.91 (m, 3H), 1.63 (m, 3H), 1.22ppm (s, 4H); MALDI-TOF MS: m/z incubatedwith3at378Cwith[3]/ACHTUNGTRENNUNG[pUC19]ratiosof0–5.0:1 overa calcdforC H N O Re:3074;found:3000[M+] 137 228 12 53 period of 24h. These solutions were then loaded into a 1% agar- Cell lines and culture conditions: An adriamycin-resistant ovarian ose gel prepared in 16 TBE buffer. Gels were then run at 70V for cancercellline(A2780/AD)andaChinesehamsterovary(CHO)cell 90min with 16 TBE as running buffer. The gel was then stained ChemMedChem2008,3,1387–1394 H2008Wiley-VCHVerlagGmbH&Co.KGaA,Weinheim www.chemmedchem.org 1393 MED R.P.Doyleetal. with ethidium bromide (10mL to 100mL TBE buffer) for 1h and K.E. O’Day, C.B. Darrow, S.M. Lane, T.A. Peyser, J.H. Satcher,Jr., W.P. rinsedwithHOfor20min. vanAntwerp,A.J.Nelson,J.G.Reynolds,Inorg.Chem.2002,41,1662– 2 1669. TopoisomeraseI inhibition assays: fX174RFI DNA was used for [11] L.Melendez,C.Decristoforo,S.J.Mather,Nucl.Med.Commun.2001,22, TopIinhibitionassays.Eachreactionmixturehadafinalvolumeof 452. 25mLthatcontained0.5mgofplasmidDNA,oneunitofTopIand [12] K.Liepe,J.Kotzerke,Nucl.Med.Commun.2007,28,623–630. varying concentrations of 3 [1–5000mm] in buffer that contained [13] a)M.J.Willhauck,B.-R.SharifSamani,F.-J.Gildehaus,I.Wolf,R.Seneko- the following: 50mm Tris–HCl pH7.5, 50mm KCl, 10mm MgCl, witsch-Schmidtke, H.-J. Stark, B. Gçke, J.C. Morris, C. Spitzweg, J. Clin. 2 0.5mmDTT,0.1mmEDTA,30mgmL(cid:1)1BSA.Solutionswereincubat- Endocrinol.Metab.2007,92,4451–4458;b)C.MQller,P.A.Schubiger,R. Schibli,Nucl.Med.Biol.2007,34,595–601. ed at 378C for 30min. The samples were then analyzed on a 1% [14] A.R.Vortherms,R.P.Doyle,NucleosidesNucleotidesNucleicAcids2008, agarose gel that was prepared in 16 TBE at 70V for 90min with 27,173–185. 1XTBE as running buffer. The gel was then stained with ethidium [15] W.Guo,R.Lee,AAPSPharmSciTech2000,1,1–7. bromide(10mL to100mLTBEbuffer)for 1h,and rinsedwithH 2 O [16] K.A.Stephenson,S.R.Banerjee,T.Besanger,O.O.Sogbein,M.K.Leva- for20min. dala, N. McFarlane, J.A. Lemon, D.R. Boreham, K.P. Maresca, J.D. Brennan, J.W. Babich, J. Zubieta, J.F. Valliant, J. Am. Chem. Soc. 2004, DNA intercalation studies: This procedure followed the protocol 126,8598–8599. of Strekas etal. with modifications.[40] To 12.5mm 3 in DMSO/H 2 O [17] C.P.Leamon,P.S.Low,Biochem.J.1993,291,855–860. (1:9), 0.3mg pUC19 DNA was added. The electron absorption was [18] A. Gabizon, A.T. Horowitz, D. Goren, D. Tzemach, F. Mandelbaum- then scanned with a UV/Vis spectrometer from 270–550nm at a Shavit,M.M.Qazen,S.Zalipsky,BioconjugateChem.1999,10,289–298. maintained temperature of 378C. The absorption was monitored [19] W.Wang,Y.K.Yan,T.S.A.Hor,J.J.Vittal,J.R.Wheaton,I.H.Hall,Poly- every 30min over 24h. The difference in initial and final absorp- hedron2002,21,1991–1999. tion readings were calculated and analyzed for peak and/or ab- [20] D.L.Ma,C.M.Che,F.M.Siu,M.Yang,K.Y.Wong,Inorg.Chem.2007,46, 740–749. sorptionshifts. [21] R.J. Lee, P.S. Low, Biochim. Biophys. Acta Biomembr. 1995, 1233, 134– Supporting Information: MS data and confocal microscopy depth 144. scanning showing images taken from layers through a single [22] M.Girasole,A.Crisenti,R.Generosi,G.Longo,G.Pompeo,S.Cotesta,A. Congiu-Castellano,Microsc.Res.Tech.2007,70,912–917. A2780/ADcellareprovided. [23] S. Ohba, M. Hiramatsu, R. Edamatsu, I. Mori, A. Mori, Neurochem. Res. 1994,19,237–241. Acknowledgements [24] M.Suwalsky,B.Ungerer,L.Quevedo,F.Aguilar,C.P.Sotomayor,J.Inorg. Biochem.1998,70,233–238. [25] F.Zobi,B.Spingler,R.Alberto,ChemBioChem2005,6,1397–1405. The authors thank the iLEARN Program and Syracuse University [26] F.Zobi,B.Spingler,T.Fox,R.Alberto,Inorg.Chem.2003,42,2818–2820. forfunding.Theauthorsaregratefultothefollowing:Oluwatayo [27] W.D.McFadyen,W.A.Denny,L.P.G.Wakelin,FEBSLett.1988,228,235– 240. F. Ikotun and Anthony R. Vortherms (Department of Chemistry, [28] C.H.Huang,S.T.Crooke,CancerRes.1985,45,3768–3773. SU) for data on cisplatin IC values, Dr. Melissa Pepling (Depart- 50 [29] K.R. Fox, D. Gauvreau, D.C. Goodwin, M.J. Waring, Biochem. J. 1980, ment of Biology, SU) for the use of the confocal microscope, and 191,729–742. Dr. Anthony Ouellette for access to the Mass Spectrometry and [30] J.Sartorius,H.J.J.Schneider,J.Chem.Soc.,PerkinTrans.21997,2319– 2327. ProteomicsCenteratSUNY,Oswego. [31] B.Jin,H.M.Lee,Y-A. Lee,J.H.Ko,C.Kim,S.K.Kim,J.Am.Chem.Soc. 2005,127,2417–2424. Keywords: confocal microscopy · cytotoxicity · DNA · folate · [32] J. Mann, A. Baron, Y. Opoku-Boahen, E. Johansson, G. Parkinson, L.R. Kelland,J.Neidle,J.Med.Chem.2001,44,138–144. rhenium [33] L. Wei, J.W. Babich, W. Ouellette, J. Zubieta, Inorg. Chem. 2006, 45, 3057–3066. [1] B.A.Kamen,M.Wang,A.J.Streckfuss,X.Peryea,R.G.Anderson,J.Biol. [34] I. Husain, J.L. Mohler, H.F. Seigler, J.M. Besterman, Cancer Res. 1994, Chem.1988,263,13602–13609. 54,539–546. [2] F. Shen, M. Wu, J. Ross, D. Miller, M. Ratnam, Biochemistry 1995, 34, [35] A.M.Codegoni,S.Castagna,C.Mangioni,A.I.Scovassi,M.Broggini,M. 5660–5665. D’Incalci,Ann.Oncol.1998,9,313–319. [3] C.P. Leamon, J.A. Reddy, Adv. Drug Delivery Rev. 2004, 56, 1127–1141. [36] Y.Nakajima,M.Satoshi,N.Kogami,T.Kawano,T.Iwai,J.Jpn.CancerRes. [4] A.Antony,Annu.Rev.Nutr.1996,16,501–521. 2001,92,1335–1341. [5] S. Wang, R.J. Lee, C.J. Mathias, M.A. Green, P.S. Low, Bioconjugate [37] R.E.McKnight,A.B.Gleason,J.A.Keyes,S.Sahabi,Bioorg.Med.Chem. Chem.1996,7,56–62. Lett.2007,17,1013–1017. [6] C.J. Mathias, S. Wang, R.J. Lee, D.J. Waters, P.S. Low, M.A. Green, J. [38] N.Lazarova,S.James,J.Babich,J.Zubieta,Inorg.Chem.Commun.2004, Nucl.Med.1996,37,1003–1008. 7,1023–1026. [7] S. Wang, J. Luo, D.A. Lantrip, D.J. Waters, C.J. Mathias, M.A. Green, [39] N.Viola-Villegas,A.R.Vortherms,R.P.Doyle,DrugTargetInsights2008, P.L.Fuchs,P.S.Low,BioconjugateChem.1997,8,673–679. 3,13–25. [8] R.Rossin,D.Pan,K.Qi,J.L.Turner,X.Sun,K.L.Wooley,M.J.Welch,J. [40] S.A.Tysoe,A.D.Baker,T.C.Strekas,J.Phys.Chem.1993,97,1707–1711. Nucl.Med.2005,46,1210–1218. [9] M. Liu, W. Xu, L.J. Xu, G.R. Zhong, S.L. Chen, W.Y. Lu, Bioconjugate Chem.2005,16,1126–1132. Received:April22,2008 [10] a)C.MQller,C.Dumas,U.Hoffmann,P.A.Schubiger,R.Schibli,J.Orga- Revised:May20,2008 nomet.Chem.2004,689,4712–4721;b)D.R.Cary,N.P.Zaitseva,K.Gray, PublishedonlineonJuly23,2008 1394 www.chemmedchem.org H2008Wiley-VCHVerlagGmbH&Co.KGaA,Weinheim ChemMedChem2008,3,1387–1394