Cancer Research Selective Visualization of Cyclooxygenase-2 in … · Therapeutics, Targets, and...

3618

Therapeutics, Targets, and Chemical Biology

Cancer

Research

Selective Visualization of Cyclooxygenase-2 in Inflammationand Cancer by Targeted Fluorescent Imaging Agents

Md. Jashim Uddin1, Brenda C. Crews1, Anna L. Blobaum1, Philip J. Kingsley1, D. Lee Gorden2, J. Oliver McIntyre2,Lynn M. Matrisian2, Kotha Subbaramaiah4, Andrew J. Dannenberg4, David W. Piston3, and Lawrence J. Marnett1

Abstract

Authors' AResearch,Vanderbilt2Departme3DepartmeAstronomTennesseeCornell Uni

Note: SupResearch O

Corresponical Center615-343-73

doi: 10.115

©2010 Am

Cancer R

Down

Effective diagnosis of inflammation and cancer by molecular imaging is challenging because of interferencefrom nonselective accumulation of the contrast agents in normal tissues. Here, we report a series of novelfluorescence imaging agents that efficiently target cyclooxygenase-2 (COX-2), which is normally absent fromcells, but is found at high levels in inflammatory lesions and in many premalignant and malignant tumors.After either i.p. or i.v. injection, these reagents become highly enriched in inflamed or tumor tissue comparedwith normal tissue and this accumulation provides sufficient signal for in vivo fluorescence imaging. Further,we show that only the intact parent compound is found in the region of interest. COX-2–specific delivery wasunambiguously confirmed using animals bearing targeted deletions of COX-2 and by blocking the COX-2active site with high-affinity inhibitors in both in vitro and in vivo models. Because of their high specificity,contrast, and detectability, these fluorocoxibs are ideal candidates for detection of inflammatory lesionsor early-stage COX-2–expressing human cancers, such as those in the esophagus, oropharynx, and colon.Cancer Res; 70(9); 3618–27. ©2010 AACR.

Introduction

Molecular imaging presents exciting opportunities for theselective detection of specific cell populations, such as thosebearing markers of disease (1, 2). Cyclooxygenase-2 (COX-2)is an attractive target for molecular imaging because it isexpressed in only a few normal tissues and is greatly upre-gulated in inflamed tissues as well as many premalignantand malignant tumors (3, 4). COX-2 is an importantcontributor to the etiology of inflammation and canceras illustrated by the efficacy of COX-2–selective inhibitorsas anti-inflammatory agents, cancer preventive agents, andadjuvant cancer therapeutic agents (5). The importance ofCOX-2 in tumor progression has been thoroughly documen-ted in the esophagus and colon where COX-2 is detected inpremalignant lesions and its levels seem to increase duringtumor progression (6–8). The importance of COX-2 insurvival and response to therapy has been elegantly shown

ffiliations: 1A.B. Hancock, Jr. Memorial Laboratory for CancerDepartments of Biochemistry, Chemistry, and Pharmacology,Institute of Chemical Biology, Center in Molecular Toxicology;nt of Cancer Biology, Vanderbilt-Ingram Cancer Center; andnts of Molecular Physiology and Biophysics and Physics andy, Vanderbilt University School of Medicine, Nashville,; and 4Department of Medicine, Weill Medical College ofversity, New York, New York

plementary data for this article are available at Cancernline (http://cancerres.aacrjournals.org/).

ding Author: Lawrence J. Marnett, Vanderbilt University Med-, 23rd Avenue and Pierce, Nashville TN, 37232-0146. Phone:29; Fax: 615-343-7534; E-mail: [email protected].

8/0008-5472.CAN-09-2664

erican Association for Cancer Research.

es; 70(9) May 1, 2010

on July 4, 2021cancerres.aacrjournals.org loaded from

by Edelman and colleagues (9) who reported that non–smallcell lung cancer patients expressing high levels of COX-2 intheir tumors have reduced survival compared with patientsexpressing low levels of COX-2. Patients with high tumor ex-pression of COX-2 benefit from the combination of carbopla-tin and gemcitabine plus the COX-2 inhibitor, celecoxib,whereas patients with low expression exhibit a poorerresponse to carboplatin/gemcitabine/celecoxib than tocarboplatin/gemcitabine alone (9).Positron emission tomography or single-photon emission

computed tomography imaging agents (18F-, 11C-, or 123I-labeled COX-2 inhibitors) have been described for nuclearimaging (10–17). These have all been based on the diaryl-heterocycle structural class analogous to celecoxib androfecoxib. Although selective uptake into macrophages ortumor cells expressing COX-2 has been shown in vitro forsome compounds, such selectivity has not been rigorouslyshown in vivo and significant nonspecific binding has beenobserved (18). Thus, despite recognition of the potential ofCOX-2–targeted imaging agents, in vivo proof-of-conceptfor this strategy is lacking.Fluorescent COX-2 inhibitors are attractive candidates as

targeted imaging agents. Such compounds have the advan-tage that each molecule bears the fluorescent tag and thecompounds are nonradioactive and stable. Thus, they canbe used conveniently for cellular imaging, animal imaging,and clinical imaging of tissues in which topical or endolumi-nal illumination is possible (e.g., esophagus, colon, and upperairway through endoscopy, colonoscopy, and bronchoscopy,respectively). Prior work from our laboratory showed thatfluorescent COX-2 inhibitors can be useful biochemicalprobes of protein binding but these earlier compounds were

. © 2010 American Association for Cancer Research.

http://cancerres.aacrjournals.org/

COX-2 Imaging

neither potent inhibitors of COX-2 nor did they possess ap-propriate fluorescence properties to be useful for cellular orin vivo imaging (19). Thus, we initiated a program to designand synthesize a series of fluorescent COX-2 inhibitors thatcould be used for these applications. The design strategy forcandidate development was based on our prior discovery thatamide derivatives of the nonselective COX inhibitor, indo-methacin, are selective COX-2 inhibitors. Many compoundswere synthesized and screened for COX-2 inhibition in vitroand in intact cells. Then, the most promising compoundswere evaluated as imaging agents in intact cells and in animalmodels of inflammation and cancer. We describe herein theoptimized candidates, their selective uptake by COX-2–expressing cells and tumors, and genetic and pharmacologicvalidation that their in vivo target is COX-2.

Materials and Methods

Synthesis and characterization of all compounds isdescribed in Supplementary Data.Inhibition assay using purified COX-1 and COX-2.

Cyclooxygenase activity of ovine COX-1 or human COX-2was assayed by a method that quantifies the conversion of[1-14C]arachidonic acid to [1-14C]prostaglandin products. Re-action mixtures of 200 μL consisted of hematin-reconstitutedprotein in 100 mmol/L Tris-HCl (pH 8.0), 500 μmol/L phenol,and [1-14C]arachidonic acid (50 μmol/L, approximately 55–57 mCi/mmol, Perkin-Elmer). For the time-dependent inhi-bition assay, hematin-reconstituted COX-1 (44 nmol/L) orCOX-2 (66 nmol/L) was preincubated at 25°C for 17 minutesand 37°C for 3 minutes with varying inhibitor concentrationsin DMSO followed by the addition of [1-14C]arachidonic acid(50 μmol/L) for 30 seconds at 37°C. Reactions were termi-nated by solvent extraction in Et2O/CH3OH/1 mol/L citrate(pH 4.0; 30:4:1). The phases were separated by centrifugationat 2,000 g for 2 minutes and the organic phase was spottedon a TLC plate (EMD Kieselgel 60, VWR). The plate wasdeveloped in EtOAc/CH2Cl2/glacial AcOH (75:25:1) at 4°C.Radiolabeled products were quantified with a radioactivityscanner (Bioscan, Inc.). The percentage of total productsobserved at different inhibitor concentrations was dividedby the percentage of products observed for protein samplespreincubated for the same time with DMSO.Cell culture and in vitro intact cell metabolism assay.

HCT116, ATCC CCL-247 human colorectal carcinoma cells,passage 8 to 18, Mycoplasma negative by a PCR detec-tion method (Sigma VenorGem) were grown in DMEM(Invitrogen/Life Technologies) + 10% fetal bovine serum(FBS; Atlas) to 70% confluence. RAW264.7, ATCC TIB-71murine macrophage-like cells, passage number 8 to 15, Myco-plasma negative by a PCR detection method were grownin DMEM + 10% heat-inactivated FBS to 40% confluence(six-well plates, Sarstedt) and activated for 7 hours in 2 mLserum-free DMEM with 200 ng/mL lipopolysaccharide (LPS;Calbiochem) and 10 μ/mL IFN γ (Calbiochem). Human 1483head and neck squamous cell carcinoma (HNSCC) cells (20),derived, characterized, and provided by Dr. Peter Sacks (NewYork University School of Dentistry, New York, NY), were

www.aacrjournals.org

on July 4, 2021cancerres.aacrjournals.org Downloaded from

grown at passage 8 to 18, Mycoplasma negative by a PCRdetection method, in DMEM/F12 + 10% FBS + Antibiotic/Antimycotic in six-well plates to 60% confluence. Serum-freemedium (2 mL) was added and the cells were treated with in-hibitor dissolved in DMSO (0–5 μmol/L, final concentration)for 30 minutes at 37°C followed by the addition of [1-14C]-arachidonic acid (10 μmol/L, ∼55 mCi/mmol) for 20 minutesat 37°C. Reactions were terminated by solvent extraction inEt2O/CH3OH/1 mol/L citrate (pH 4.0; 30:4:1) and the organicphase was spotted on a 20 × 20 cm TLC plate (EMD Kieselgel60, VWR). The plate was developed in EtOAc/CH2Cl2/glacialAcOH (75:25:1) and radiolabeled products were quantified witha radioactivity scanner (Bioscan, Inc.). The percentage of totalproducts observed at different inhibitor concentrations wasdivided by the percentage of products observed for cellspreincubated with DMSO.Fluorescence microscopy of RAW264.7 cells or 1483

HNSCC cells. RAW264.7 cells were plated on 35-mm MatTekdishes (MatTek Corp.) such that the cells were 40% confluentand human 1483 HNSCC cells were 60% confluent on the dayof the experiment. The RAW264.7 cells were activated for6 hours in serum-free DMEM with 200 ng/mL LPS and10 μ/mL IFN γ. Both cell lines were incubated in 2.0 mLHBSS/Tyrode's buffer with 200 nmol/L compound 1, 2, or 3for 30 minutes at 37°C. To block the COX-2 active site, thecells were preincubated with 10 μmol/L indomethacin or5 μmol/L celecoxib for 20 minutes before the addition ofcompound 1 or 2. The cells were then washed brieflythrice and incubated in HBSS/Tyrode's buffer for 30 minutesat 37°C. Following the required washout period, the cellswere imaged in 2.0 mL fresh HBSS/Tyrode's buffer on aZeiss Axiovert 25 Microscope with the propidium iodidefilter (0.5–1.0 second exposure, gain of 2). All treatmentswere performed in duplicate dishes in at least three separateexperiments.Confocal microscopy of 1483 cells treated with

compound 2/mitotrackerGR. 1483 HNSCC were plated inMatTek dishes (MatTek Corp.) and grown to 60% to 70% con-fluence for 48 hours. DMSO or compound 2 (100 nmol/L) wasadded to each dish containing 2.0mLHBSS/Tyrode's buffer for30 minutes at 37°C. After four quick HBSS washes, cells wereincubated for 30 minutes in 2.0 mL HBSS/Tyrode's buffer andimaged with a Zeiss LSM510 confocal microscope using a 63 ×1.4 NA plan-Apochromat objective lens. To visualize cellularmitochondria, 100 nmol/L Mitotracker GR was added for15 minutes at 37°C followed by three quick washes before im-aging. Four hundred eighty-eight nanometer excitation wereused to image Mitotracker GR through a 500- to 530-nm band-pass filter and compound 2 was imaged using 532 nm excita-tion and collection through a 565- to 615-nm bandpass filter.The pinhole was set to 1 Airy unit and images were collectedthroughout the focus of the cells. To assure a full sampling ofthe perinuclear region, analysis was performed on the opticalsections through the middle of the nucleus.In vivo imaging of COX-2 in inflammation. Carrageenan

(50 μL 1% in sterile saline) was injected in the rear leftfootpad of female C57BL/6 mice, followed by compound1 or 2 (1 mg/kg, i.p.) at 24 hours postcarrageenan. Animals

Cancer Res; 70(9) May 1, 2010 3619



Uddin et al.

3620

were imaged 3 hours later in a Xenogen IVIS 200 (DsRed fil-ter, 1.5 cm depth, 1 s). For comparison, animals also weredosed with compound 3, which does not inhibit COX-2. To testfurther the molecular target for compound 2, parallel experi-ments were performed using COX-2 (−/−) mice. Experimentsalso were performed in which compound 1 was administeredto the same animals by repetitive i.p. injection on days 1, 3, 5,and 7 to monitor the time course of compound uptake follow-ing carrageenan induction of inflammation.Establishment of xenografts in nude mice. Female nude

mice, NU-Fox1nu, were purchased at 6 to 7 weeks of agefrom Charles River Laboratories. Human 1483 HNSCC cellsand HCT116 colorectal carcinoma cells were trypsinizedand resuspended in cold PBS containing 30% Matrigel suchthat 1 × 106 cells in 100 μL were injected s.c. on the left flank.The HCT116 or 1483 xenografts required only 2 to 3 weeks ofgrowth.In vivo imaging of nude mice with xenografts. Female

nude mice bearing medium-sized HCT116 or 1483 xenografttumors on the left flank were dosed by i.p. injection with2 mg/kg compound 2 or by retro-orbital injection with1 mg/kg compound 2. The animals were lightly anesthetizedwith 2% isoflurane for fluorescence imaging in the XenogenIVIS 200 with the DSRed filter at 1.5-cm depth and 1-secondexposure (f2). For the COX-2 active site–blocking experi-ments, nude mice bearing 1483 xenografts were predosedby i.p. injection with 2 mg/kg indomethacin at 24 hoursand 1 hour before dosing with compound 2 (2 mg/kg, i.p.).Pharmacokinetics of candidate compounds. Female

nude mice with medium-sized 1483 HNSCC xenografttumors on the left flank were injected i.p. with 2 mg/kg com-pound 2. At 0, 0.5, 3, 12, and 24 hours, the mice (n = 4 for eachtime point, duplicate experiments) were anesthetized withisoflurane. Blood samples were taken by cardiac punctureinto a heparinized syringe into a 1.5-mL heparinized tubeon ice, followed by removal of the liver, kidney, contralateralleg muscle, and xenograft tumor. All organs/tissues wererinsed briefly in ice-cold PBS, blotted dry, weighed, andsnap frozen in liquid nitrogen. The blood samples werecentrifuged at 4°C at 6,000 rpm for 5 minutes and theplasma was transferred to clean tubes and frozen at −80°C.Compound 2 was extracted by homogenizing the tissuein 100 mmol/L Tris (pH 7.0) buffer and mixing an ali-quot of the homogenate with 1.2× volume of acetonitrile.The acetonitrile was removed and the samples were dried,reconstituted, and analyzed through reversed-phase high-performance liquid chromatography (HPLC)-UV using aPhenomenex 10 × 0.2 cm C18 or a Phenomenex 7.5 × 0.2 cmSynergi Hydro-RP column held at 40°C. The samples werequantified against a standard curve prepared by addingcompound 2 to tissue homogenates of undosed animalsfollowed by the workup described. Cochromatography wasperformed with multiple columns and elution conditions asdescribed in the Supplemental Data.In vivo imaging of Min mice. C57BL/6 APC-Min mice

maintained on a high-fat (11%) diet for 18 weeks developed20 to 30 intestinal polyps per mouse. Before imaging, Minmice were anesthetized (2% inhaled isoflurane) for retro-

Cancer Res; 70(9) May 1, 2010


orbital injection of compound 2 at 1 mg/kg. At 2 hours post-injection, the mice were euthanized and the intestines wereresected, washed with PBS, and fixed in 10% formalin beforeex vivo imaging by fluorescence dissecting microscopy (ZeissM2Bio; n = 5).

Results

COX-2 is a potentially ideal target for molecular imagingbecause its active site (and the active site of COX-1) is burieddeep inside each subunit of the homodimeric protein (21–23).Access to the active site is controlled by a constrictionthat separates it from a large opening in the membrane-binding domain that we have termed the lobby (Supple-mentary Fig. S1). All substrates or inhibitors bind in the lobbyand then diffuse through the constriction into the active site(24). The constriction is composed of Tyr-355, Glu-524, andArg-120 and serves as the binding site for the carboxylic acidgroup of substrates and certain inhibitors (25). We have re-ported that neutral derivatives (esters and amides) of certaincarboxylic acid inhibitors (e.g., indomethacin) bind to COX-2but not to COX-1 (26). A three-dimensional structure ofCOX-2 complexed to such a conjugate has not been solvedbut structures of related complexes suggest the indomethacinunit binds in the active site with the tethered amide, breech-ing the constriction and projecting into the lobby (22, 27).These structural and functional analyses provide the designprinciples for the construction of COX-2–targeted imagingagents.Synthesis of candidate compounds and cellular imaging.

Three carboxylic acid cores—i.e., indomethacin, a celecoxibcarboxylic acid derivative, and an indolyl carboxamide ana-logue of indomethacin—were tethered through a series ofalkylenediamines, piperazines, polyethylene glycol, or pheny-lenediamines to a diverse range of fluorophores. The fluo-rophores attached included dansyl, dabsyl, coumarin,fluorescein, rhodamine, alexa-fluor, nile blue, cy5, cy7, nearIR, and IR dyes as well as lanthanide chelators. Nearly200 compounds were synthesized and each conjugate wastested for its ability to selectively inhibit COX-2 in assaysusing purified proteins in vitro. Promising molecules weretested for their ability to inhibit COX-2 in LPS-treatedRAW264.7 macrophages. Preliminary experiments indicatedthat indomethacin conjugates bound most tightly and selec-tively to COX-2; therefore, most of the compounds synthe-sized were derived from this core.Indomethacin conjugates to dansyl, dabsyl, coumarin,

fluorescein, and rhodamine-derived fluorophores exhibitedpromising COX-2 inhibition and selectivity both in vitroand in intact cells. The carboxy-X-rhodamine (6-ROX- and5-ROX)–based conjugates, 1 and 2, displayed the best ba-lance of cellular activity and optical properties (λex = 581nm, λemit = 603 nm) and were used for all subsequent experi-ments (Table 1). A detailed kinetic analysis indicated that1 and 2 require lengthy preincubations with COX-2 toachieve maximal inhibition but once bound, they dissociatevery slowly (Supplementary Fig. S2). Thus, they are slow,tight-binding inhibitors with very low rates of association

Cancer Research



COX-2 Imaging

ww

Table 1. Biochemical properties of carboxy-X-rhodamine derivatives

Compound no.

w.aacrjournals.org

cancerrDownloaded from

Structure

on July 4, 2021. © 2010 Ames.aacrjournals.org

Purified enzyme IC50(μmol/L)

Cancer Res; 70(9)

erican Association for Cancer Re

Cell IC50(μmol/L)

COX-1
COX-2
1
>25 0.83 0.36
2
>25 0.7 0.31
3
>25 >4 ND
4
>25 >4 ND
5
>25 >4 ND
6
0.92 0.21 NT
7
>25 0.14 NT
8
0.05 0.75 0.01
NOTE: Assays were conducted as described in Materials and Methods. IC50s for inhibition of ovine COX-1 or human COX-2.Compounds also were tested in intact RAW264.7 macrophages (Cell IC50).Abbreviations: ND, no inhibition detected up to 5 μmol/L; NT, not tested.

May 1, 2010 3621

search.


Uddin et al.

3622

and dissociation. Compounds 1 and 2 were less potentthan celecoxib or rofecoxib as inhibitors of COX-2 (Table 1).A negative control molecule (3) was synthesized thatcontained 6-ROX bound to indomethacin through a shorterethylenediamine tether, which eliminated COX-2 inhibition(Table 1).The human head and neck cancer cell line, 1483, which

expresses high levels of COX-2 (28), exhibited strong labelingwith compounds 1 or 2 (Fig. 1A). Preincubation of the cellswith the COX-2–selective inhibitor, celecoxib, prevented thelabeling of 1483 cells by either compound (Fig. 1B). In all ofthese in vitro experiments, the labeling seemed to be intracel-lular, so confocal microscopy was performed to verify the lo-calization. Incubation of compound 2 with 1483 cells led tothe perinuclear labeling of membraneous structures that ap-peared to be endoplasmic reticulum or Golgi (Fig. 1C). Theperinuclear labeling correlated well to multiple previous re-ports of the intracellular localization of COX-2 (29–32). Incu-bation of the same cells with Mitotracker showed that themitochondria did not colocalize with compound 2 (Fig. 1C).The mouse macrophage–like cell line, RAW 264.7, does not

express COX-2 and exhibited very weak labeling with 1 or 2(e.g., Supplementary Fig. S3A), whereas LPS-pretreated cellslabeled more strongly (Supplementary Fig. S3B). The labelingof the COX-2–expressing RAW cells by 1 or 2 was preventedby pretreatment of the cells with indomethacin (Supplemen-tary Fig. S3C) or celecoxib, which block the COX-2 active site.Importantly, no labeling was observed when either control orLPS-pretreated RAW cells were incubated with compound 3,which does not inhibit COX-2 (Supplementary Fig. S3D). Theextent of compound 2 uptake increased at 4 hours with theappearance of COX-2 protein. A further increase in uptakewas not observed at 7 hours although there was higherCOX-2 protein as detected by Western blotting (Supplemen-tary Fig. S4). Comparison of the amount of compound 2taken up at 7 hours to the amount of COX-2 estimated byWestern blotting in the LPS-treated cells suggested a stoichi-ometry of binding of 0.90 (Supplementary Data).Imaging carrageenan–induced inflammation. Com-

pounds 1 and 2 seemed promising based on these in vitroimaging experiments, so their potential for in vivo imagingwas evaluated using carrageenan-induced inflammation inthe mouse footpad, human tumor xenografts in nude mice,and spontaneous tumors arising in mouse models. Themouse footpad model is well documented for the role ofCOX-2–derived prostaglandins as a major driving force forthe acute edema that results 24 hours after carrageenan in-jection into the paw (33). One of the significant advantages ofthis animal model of inflammation is the ability to image theinflamed footpad compared with the noninflamed contra-lateral footpad, which does not express COX-2. We injectedfemale C57BL/6 mice with 50 μL 1% carrageenan in the rearleft footpad, followed by compound 1 or 2 (1 mg/kg, i.p.)at 24 hours postcarrageenan. Animals were imaged 3 hourslater in a Xenogen IVIS 200 (DsRed filter, 1.5 cm depth, 1 s).Both compounds 1 and 2 targeted the swollen footpad withan average 4.5-fold increase in fluorescence over that of thecontralateral, uninjected footpad (Fig. 2). For comparison,



Figure 1. Labeling of COX-2–expressing cells by compound 2. Theexperimental protocols are described inMaterials andMethods. A, 1483HNSCCcells treated with 200 nmol/L compound 2 for 30 min. B, 1483 HNSCCcells pretreated with 5 μmol/L celecoxib for 20 min before compound 2treatment. C, confocal microscopy of 1483 HNSCC cells treated with bothmitotrackerGR (blue; mitochondria) and compound 2 (red; perinuclear).

Cancer Research



COX-2 Imaging

animals also were dosed with compound 3, which doesnot inhibit COX-2. In Fig. 2A, the left mouse was dosed withcompound 3 and the right mouse with compound 1. Com-pound 3 yielded minimal fluorescence in the inflamed pawcompared with the contralateral paw, whereas compound 1yielded a strong signal in the inflamed paw. To test furtherthe molecular target of compounds 1 or 2, parallel expe-riments were performed using mice bearing targeted dele-tions in COX-2. Figure 2B depicts the fold difference in thecompound 2–derived fluorescence signal in the 24-hourcarrageenan-injected footpad over the control footpad forwild-type versus COX-2 (−/−) mice. The COX-2 null miceconsistently showed approximately a 40% increase in signalin the swollen footpad apparently due to nonspecific binding.This contrasts with a 400% to 600% increase in the swollenfootpad in wild-type mice. Finally, experiments were con-ducted to evaluate the uptake of compound 1 during the res-olution of inflammation. Following carrageenan injection,compound 1 was administered i.p. 1, 3, 5, and 7 days laterand the animals were imaged. Uptake of compound 1 wasmaximal at 24 hours but declined thereafter (Fig. 2C). At-tempts to estimate active COX-2 protein by the quantificationof prostaglandins in paw extracts were unsuccessful becauseof poor recovery.Imaging COX-2–expressing tumors. The results in the

footpad inflammation model show that COX-2–targetedfluorescent conjugates are taken up in inflamed paws ofCOX-2–expressing mice but not in COX-2 null animals. Wenext evaluated the ability of these compounds to targetCOX-2 in human tumor xenografts. Female nude mice wereinjected in the left flank with HCT-116 or 1483 cells and thexenografts were allowed to grow to approximately 750 to1,000 mm3. Animals were dosed by retro-orbital injectionwith compound 2 (1 mg/kg) then lightly anesthetized with2% isoflurane in preparation for imaging. No fluorescencewas observed during the first 60 minutes postinjection, butsignal was reproducibly detected in the COX-2–expressing1483 tumors starting at 3 to 5 hours and persisted as longas 26 hours postinjection. At 3.5 hours postinjection, theHCT116 tumor, which does not express COX-2 (34),showed minimal fluorescence (Fig. 3A) whereas the 1483tumor exhibited bright fluorescence (Fig. 3B). In anothercontrol experiment, nude mice bearing 1483 xenograftswere treated with the fluorophore alone, 5-ROX (2 mg/kg, i.p.), which is neither an inhibitor of COX-2 norCOX-1. No signal from 5-ROX alone accumulated in the tu-mors at any time point. This result showed that the fluoro-phore moiety was not responsible for the tumor uptake ofcompound 2, supporting the conclusion that the difference inlabeling of 1483 and HCT116 xenografts is due their differen-tial in COX-2 expression.Nude mice with 1483 xenografts were pretreated with

either DMSO or indomethacin in DMSO (2 mg/kg, i.p.) beforecompound 2 dosing (2 mg/kg, i.p.). At 3 hours postinjection,the DMSO-pretreated mice showed strong fluorescence intheir tumors (Fig. 3C) compared with weak signals in thetumors of the indomethacin-pretreated mice (Fig. 3D). Inthe mouse xenograft model, indomethacin was able to block



Figure 2. In vivo labeling of COX-2–expression in inflammation bycompound 1, 2, or 3. A, C57BL/6 mouse with carrageenan-inducedinflammation in the left foot pad. The left mouse was dosed with thenegative control molecule 3 (1 mg/kg, i.p.) and the right mouse wasdosed with compound 1 (1 mg/kg, i.p.). Both mice were imaged at3 h postinjection. B, fold increase of fluorescence in inflamed versuscontralateral paw of wild-type (WT) and COX-2 (−/−) mice at 3 hpostinjection of compound 2 (1 mg/kg, i.p.; n = 6). RFU, relativefluorescence unit. C, carrageenan was injected in the rear left footpads offemale C57BL/6 mice, followed by dosing compound 1 (1 mg/kg i.p.)24 h later. Animals were reinjected with compound 1 at 3, 5, and7 d postcarrageenan (n = 9). Mice were imaged at 3 h aftercompound injection. The plot shows the fold increase of fluorescence inswollen versus contralateral foot (n = 6).

Cancer Res; 70(9) May 1, 2010 3623



Uddin et al.

3624

92 ± 6% (n = 8) of the COX-2–expressing tumor uptake ofcompound 2.We next investigated whether the COX-2 inhibitory activity

of our imaging probes correlated with their in vivo efficacyin targeting COX-2–expressing tumors. Nude mice bearing1483 xenografts were dosed (2 mg/kg, i.p.) with compound 4(no COX inhibition at 3 μmol/L), compound 5 (30% COX-2inhibition at 3 μmol/L), and compound 2 (90% COX-2 inhi-bition at 3 μmol/L; Table 1). At 3.5 hours postinjection, fluo-rescence from the tumor region was directly proportionalto the compound potency as a COX-2 inhibitor (Supplemen-tary Fig. S5).Experiments were conducted to determine the identity of

the fluorescent material(s) detected in vivo and to monitorthe time course of its distribution and tissue uptake followinginjection of compound 2 into nude mice bearing 1483 humantumor xenografts. Extracts of plasma, liver, kidney, tumor,and adjacent muscle were quantitatively analyzed by HPLCat different times after i.p. administration of the compound.A single fluorescent compound was detected in all the ex-tracts, which coeluted with a standard of compound 2 inmultiple HPLC systems. This compound displayed an identi-cal mass spectrum to the unmetabolized parent molecule,compound 2 (Fig. 4A). The time courses of uptake and dis-tribution of compound 2 in plasma and various tissues aredisplayed in Fig. 4B. Compound 2 was rapidly distributed fol-lowing i.p. administration and reached nearly maximal levelsin plasma, liver, and kidney 30 minutes after injection. Com-pound 2 levels declined substantially over the course of 12 to24 hours to a small fraction of its initial levels in all three ofthese compartments. In contrast, the time course for uptakeof compound 2 into the 1483 tumors lagged substantially andrequired ∼3 hours to reach near maximal levels. The levels ofcompound 2 remained relatively high in the tumor so by24 hours, the tumor levels were as high as the levels in liveror kidney. This indicates both slow uptake and release ofcompound 2 into and out of the tumor.APCmin mice bear the same mutation (Apc−) that is caus-

ative for familial adenomatous polyposis in human beingsand these mice primarily develop small intestinal tumorsthat express COX-2 (35, 36). Crossing APCmin mice withCOX-2 (−/−) mice reduces intestinal tumor development by85% and treatment of APCmin mice with COX-2 inhibitorsalso reduces tumorigenesis (37, 38). APCmin mice (ages 18–20 wk, fed a high-fat diet) were injected retro-orbitally withcompound 2 (1 mg/kg), and after 2 hours, animals were sac-rificed and their intestines were removed. The tissue waswashed thoroughly with PBS, opened longitudinally, andimaged. Figure 5A shows the low background fluorescenceof a section of small intestine without polyps. A single polyp(Fig. 5B) and a five-polyp cluster (Fig. 5C) displayed highfluorescence, with greatly increased detection compared withbright-field visualization. The signal enrichment of com-pound 2 in the polyps was estimated to be >50:1. COX-2 ex-pression in the polyps seems to be required for this selectiveuptake although other factors beside the level of COX-2protein may contribute to the relative enrichment over sur-rounding normal tissue.



Discussion

These studies show the feasibility of specific in vivo target-ing of COX-2 in inflammatory lesions and tumors usingorganic fluorophores tethered to indomethacin through anamide linkage. Compounds 1 and 2 display a very high degreeof selectivity of uptake by inflammatory tissue and tumors inlive animals relative to surrounding normal tissue or muscleas determined by either imaging or mass spectrometry. Thisselectivity seems greater than that reported in previous liter-ature reports of fluorescent tumor imaging in which the ratioof tumor fluourescence was compared with muscle fluores-cence (39). Uptake of our compounds requires the expression

Figure 3. In vivo labeling of COX-2–expressing xenografts bycompound 2. A, nude mice with HCT116 xenograft (COX-2 negative)or 1483 xenograft (B; COX-2 positive) were dosed (retro-orbital) with1 mg/kg compound 2 and imaged at 3.5 h postinjection. C, nude micewith 1483 xenografts were predosed with DMSO before injection ofcompound 2 (2 mg/kg, i.p.), or predosed with indomethacin (D; 2 mg/kg,i.p.) 24 and 1 h before compound 2 and imaged at 3 h postinjection(Xenogen IVIS, DsRed filter, 1 s, f2, 1.5 cm depth). The emission observedaround the peritoneal cavity in C and D is due to residual compound2 at the site of injection.

Cancer Research



COX-2 Imaging

of COX-2 at the target site and declines as the level of COX-2decreases. Although uptake into inflamed or tumor tissueseems to be slower than expected from simple distributionin the body, the kinetics of compound release seem to beextremely slow, thus leading to a detectable buildup of thelabel. Similar results are observed by both imaging com-pounds 1 or 2 (Figs. 2 and 3) or by direct quantitative analysisof compound 2 (Supplementary Fig. S4; Fig. 4).To achieve this success, ~200 compounds were evaluated

as candidate COX-2–targeted imaging agents. Although a sig-



nificant percentage showed COX-2 inhibitory activity againstpurified protein, only a fraction of these compounds inhib-ited COX-2 activity in intact cells, and of those, most didnot possess fluorescence properties suitable for in vivo imag-ing. Among the compounds that emerged from our develop-ment pathway, only compounds 1 and 2 exhibited sufficientmetabolic stability to survive long enough to distribute toinflammatory lesions or xenograft tumors. The low overallsuccess rate (∼1%) likely underscores why COX-2–targetedimaging agents have proven difficult to develop.

Figure 4. Analysis of fluorescentmaterial in xenografts and severalmouse tissues. A, representativeHPLC-UV chromatogram(detection, 581 nm) of1483 tumor extract (4 hpostadministration) revealed asingle major fluorescentcompound that coeluted withcompound 2 (15.3 min). Thepeak eluting at 13.3 minintegrates for

Uddin et al.

3626

The specificity for COX-2 binding of these compounds wasillustrated by multiple observations: (a) only cultured cellsthat express COX-2 took up fluorocoxibs and uptake was in-hibited by the COX inhibitors indomethacin and celecoxib.



The intracellular localization of the probes matches that ofCOX-2 protein and the stoichiometry of uptake was ∼0.9molecule of beacon per subunit; (b) uptake into inflamedover noninflamed tissue was blocked by indomethacinpretreatment of the animals and was not observed inCOX-2 (−/−) animals. No uptake was observed with aclose structural analogue of compound 1 that does not in-hibit COX-2; (c) uptake into COX-2–expressing tumors wasblocked by indomethacin pretreatment of the animals anda correlation was found between the amount of light emis-sion from the tumor and the COX-2 inhibitory potencyof the beacon. The nontargeted fluorophores 5-ROX and6-ROX did not accumulate in COX-2–expressing xenografts;and (d) >95% of the fluorescent material present in the tu-mors is the unmetabolized parent compound. Thus, in vitroand in vivo studies provide strong support for the conclusionthat binding to COX-2 is the major determinant of uptake intoinflamed, premalignant, or malignant tissue. Although thestoichiometry of compound 2 binding to COX-2 protein wasestimated to be 0.9 in activated RAW cells, such high stoichi-ometry cannot be assumed in all situations. The extent ofuptake in cells, inflamed tissue, or tumors will depend uponseveral factors such as the permeability of COX-2–expressingcells to the probe, kinetics of binding and release from theCOX-2 active site, vascularization of the tissue, and possibleexpulsion of the probes by transporters. Further studies willbe needed to explore in greater detail the quantitative aspectsof the use of these compounds in vivo.Compounds 1 and 2 represent the first feasible reagents

for clinical detection of tissues containing high levels ofCOX-2 in settings amenable to fluorescent excitation andanalysis by surface measurement or endoscopy (e.g., skin,esophagus, intestine, and bladder). Although such fluoroco-xibs will not be useful for applications in internal organs thatare not accessible for optical imaging, their developmentprovides rigorous proof of concept for the feasibility ofmolecular targeting of COX-2 in inflammatory lesions,premalignant lesions, and tumors.

Disclosure of Potential Conflicts of Interest

L.J. Marnett: sponsored research and consulting, XL Tech Group. The otherauthors disclosed no potential conflicts of interest.

Acknowledgments

We thank S.K. Dey for the COX-2 null animals and Melissa Turman for theassistance with molecular graphics.

Grant Support

Research and Center grants from the NIH (CA86283, CA89450, CA105296,CA68485, CA60867, CA126588, CA111469, and GM72048), the Medical Free-Electron Laser Program of the US Department of Defense, XL TechGroup,and New York Crohn's Foundation.

The costs of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 07/16/2009; revised 02/20/2010; accepted 02/23/2010; publishedonline 04/29/2010.

Figure 5. In vivo labeling of COX-2–expression in intestinal polyps bycompound 2. C57BL/6J-Min/+ mice bearing small intestinal polypswere euthanized at 2 h after retro-orbital injection of compound 2(1 mg/kg) and small intestines were washed, opened, and examined bydissecting fluorescence microscopy. A, section of small intestine withno polyp, 90-millisecond exposure. B, single polyp, 90-millisecondexposure. C, polyp cluster, 90-millisecond exposure.

Cancer Research



COX-2 Imaging

References

1. Weissleder R, Ntziachristos V. Shedding light onto live molecular

targets. Nat Med 2003;9:123–8.2. Tanabe K, Zhang Z, Ito T, Hatta H, Nishimoto S. Current molecular

design of intelligent drugs and imaging probes targeting tumor-specific microenvironments. Org Biomol Chem 2007;5:3745–57.

3. Crofford LJ. COX-1 and COX-2 tissue expression: implications andpredictions. The J Rheumatol 1997;49:15–9.

4. Dannenberg AJ, Lippman SM, Mann JR, Subbaramaiah K, DuBois RN.Cyclooxygenase-2 and epidermal growth factor receptor:pharmacologic targets for chemoprevention. J Clin Oncol 2005;23:254–66.

5. Marnett LJ. The COXIB experience: a look in the rear-view mirror.Annu Rev Pharmacol Toxicol 2009;49:265–90.

6. Kandil HM, Tanner G, Smalley W, Halter S, Radhika A, Dubois RN.Cyclooxygenase-2 expression in Barrett's esophagus. Dig Dis Sci2001;46:785–9.

7. Eberhart CE, Coffey RJ, Radhika A, Giardiello FM, Ferrenbach S,DuBois RN. Up-regulation of cyclooxygenase 2 gene expressionin human colorectal adenomas and adenocarcinomas. Gastroen-terology 1994;107:1183–8.

8. Kargman SL, O'Neill GP, Vickers PJ, Evans JF, Mancini JA, Jothy S.Expression of prostaglandin G/H synthase-1 and -2 protein inhuman colon cancer. Cancer Res 1995;55:2556–9.

9. Edelman MJ, Watson D, Wang X, et al. Eicosanoid modulation inadvanced lung cancer: cyclooxygenase-2 expression is a positivepredictive factor for celecoxib + chemotherapy-Cancer andLeukemia Group B Trial 30203. J Clin Oncol 2008;26:848–55.

10. Toyokuni T, Kumar JS, Walsh JC, et al. Synthesis of 4-(5-[18F]fluoro-methyl-3-phenylisoxazol-4-yl)benzenesulfonamide, a new [18F]fluo-rinated analogue of valdecoxib, as a potential radiotracer forimaging cyclooxygenase-2 with positron emission tomography.Bioorg Med Chem Lett 2005;15:4699–702.

11. Tanaka M, Fujisaki Y, Kawamura K, et al. Radiosynthesis andevaluation of 11C-labeled diaryl-substituted imidazole and indolederivatives for mapping cyclooxygenase-2. Biol Pharm Bull 2006;29:2087–94.

12. Majo VJ, Prabhakaran J, Simpson NR, Van Heertum RL, Mann JJ,Kumar JS. A general method for the synthesis of aryl [11C]methyl-sulfones: potential PET probes for imaging cyclooxygenase-2 ex-pression. Bioorg Med Chem Lett 2005;15:4268–71.

13. Prabhakaran J, Underwood MD, Parsey RV, et al. Synthesis andin vivo evaluation of [18F]-4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benzenesulfonamide as a PET imaging probe forCOX-2 expression. Bioorg Med Chem 2007;15:1802–7.

14. Wuest F, Kniess T, Bergmann R, Pietzsch J. Synthesis andevaluation in vitro and in vivo of a 11C-labeled cyclooxygenase-2(COX-2) inhibitor. Bioorg Med Chem 2008;16:7662–70.

15. Schuller HM, Kabalka G, Smith G, Mereddy A, Akula M, Cekanova M.Detection of overexpressed COX-2 in precancerous lesions ofhamster pancreas and lungs by molecular imaging: implications forearly diagnosis and prevention. Chem Med Chem 2006;1:603–10.

16. de Vries EF, van Waarde A, Buursma AR, Vaalburg W. Synthesisand in vivo evaluation of 18F-desbromo-DuP-697 as a PET tracerfor cyclooxygenase-2 expression. J Nucl Med 2003;44:1700–6.

17. de Vries EF, Doorduin J, Dierckx RA, van Waarde A. Evaluation of[11C]rofecoxib as PET tracer for cyclooxygenase 2 overexpressionin rat models of inflammation. Nucl Med Biol 2008;35:35–42.

18. de Vries EF. Imaging of cyclooxygenase-2 (COX-2) expression:potential use in diagnosis and drug evaluation. Curr Pharm Des2006;12:3847–56.

19. Timofeevski SL, Prusakiewicz JJ, Rouzer CA, Marnett LJ. Isoform-selective interaction of cyclooxygenase-2 with indomethacinamides studied by real-time fluorescence, inhibition kinetics, andsite-directed mutagenesis. Biochemistry 2002;41:9654–62.

20. Sacks PG, Parnes SM, Gallick GE, et al. Establishment andcharacterization of two new squamous cell carcinoma cell lines



derived from tumors of the head and neck. Cancer Res 1988;48:2858–66.

21. Picot D, Loll PJ, Garavito RM. The X-ray crystal structure of themembrane protein prostaglandin H2 synthase-1. Nature 1994;367:243–9.

22. Luong C, Miller A, Barnett J, Chow J, Ramesha C, Browner MF.Flexibility of the NSAID binding site in the structure of humancyclooxygenase-2. Nature Struct Biol 1996;3:927–33.

23. Kurumbail RG, Stevens AM, Gierse JK, et al. Structural basis forselective inhibition of cyclooxygenase-2 by anti-inflammatoryagents. Nature 1996;384:644–8.

24. Kurumbail RG, Kiefer JR, Marnett LJ. Cyclooxygenase enzymes:catalysis and inhibition. Curr Opin Struct Biol 2001;11:752–60.

25. Bhattacharyya DK, Lecomte M, Rieke CJ, Garavito RM, Smith WL.Involvement of arginine 120, glutamate 524, and tyrosine 355 inthe binding of arachidonate and 2-phenylpropionic acid inhibitorsto the cyclooxygenase active site of ovine prostaglandin endo-peroxide H synthase-1. J Biol Chem 1996;271:2179–84.

26. Kalgutkar AS, Crews BC, Rowlinson SW, et al. Biochemically baseddesign of cyclooxygenase-2 (COX-2) inhibitors: facile conversion ofnonsteroidal antiiflammatory drugs to potent and highly selectiveCOX-2 inhibitors. Proc Natl Acad Sci USA 2000;97:925–30.

27. Harman CA, Turman MV, Kozak KR, Marnett LJ, Smith WL,Garavito RM. Structural basis of enantioselective inhibition ofcyclooxygenase-1 by S-α-substituted indomethacin ethanolamides.J Biol Chem 2007;282:28096–105.

28. Zweifel BS, Davis TW, Ornberg RL, Masferrer JL. Direct evidencefor a role of cyclooxygenase 2-derived prostaglandin E2 in humanhead and neck xenograft tumors. Cancer Res 2002;62:6706–11.

29. Regier MK, DeWitt DL, Schindler MS, Smith WL. Subcellularlocalization of prostaglandin endoperoxide synthase-2 in murine3T3 cells. Arch Biochem Biophys 1993;301:439–44.

30. Coffey RJ, Hawkey CJ, Damstrup L, et al. Epidermal growth factorreceptor activation induces nuclear targeting of cyclooxygenase-2,basolaterial release of prostaglandins, and mitogenesis in polarizingcolon cancer cells. Proc Natl Acad Sci USA 1997;94:657–62.

31. Spencer AG, Woods JW, Arakawa T, Singer II, Smith WL. Subcellularlocalization of prostaglandin endoperoxide H synthases-1 and -2 byimmunoelectron microscopy. J Biol Chem 1998;273:9886–93.

32. Mbonye UR, Wada M, Rieke CJ, Tang HY, Dewitt DL, Smith WL.The 19-amino acid cassette of cyclooxygenase-2 mediates entryof the protein into the endoplasmic reticulum-associated degrada-tion system. J Biol Chem 2006;281:35770–8.

33. di Meglio P, Ianaro A, Ghosh S. Amelioration of acute inflammationby systemic administration of a cell-permeable peptide inhibitor ofNF-κB activation. Arthritis and rheumatism 2005;52:951–8.

34. Sheng HM, Shao JY, Kirkland SC, et al. Inhibition of human coloncancer cell growth by selective inhibition of cyclooxygenase-2. J ClinInvest 1997;99:2254–9.

35. Moser AR, Pitot HC, Dove WF. A dominant mutation that pre-disposes to multiple intestinal neoplasia in the mouse. Science1990;247:322–4.

36. Williams CS, Luongo C, Radhika A, et al. Elevated cyclooxygenase-2levels in Min mouse adenomas. Gastroenterology 1996;111:1134–40.

37. Oshima M, Dinchuk JE, Kargman S, et al. Suppression of intestinalpolyposis in ApcΔ716 knockout mice by inhibition of cyclooxygenase2 (COX-2). Cell 1996;87:803–9.

38. Jacoby RF, Seibert K, Cole CE, Kelloff G, Lubet RA. The cyclooxy-genase-2 inhibitor celecoxib is a potent preventive and therapeuticagent in the min mouse model of adenomatous polyposis. CancerRes 2000;60:5040–4.

39. Bugaj JE, Achilefu S, Dorshow RB, Rajagopalan R. Novel fluorescentcontrast agents for optical imaging of in vivo tumors based on areceptor-targeted dye-peptide conjugate platform. J Biomed Opt2001;6:122–33.

Cancer Res; 70(9) May 1, 2010 3627



2010;70:3618-3627. Cancer Res Md. Jashim Uddin, Brenda C. Crews, Anna L. Blobaum, et al. and Cancer by Targeted Fluorescent Imaging AgentsSelective Visualization of Cyclooxygenase-2 in Inflammation

Updated version

http://cancerres.aacrjournals.org/content/70/9/3618

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2010/05/10/70.9.3618.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/70/9/3618.full#ref-list-1

This article cites 39 articles, 14 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/70/9/3618.full#related-urls

This article has been cited by 8 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/70/9/3618To request permission to re-use all or part of this article, use this link

on July 4, 2021. © 2010 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
http://cancerres.aacrjournals.org/content/70/9/3618http://cancerres.aacrjournals.org/content/suppl/2010/05/10/70.9.3618.DC1http://cancerres.aacrjournals.org/content/70/9/3618.full#ref-list-1http://cancerres.aacrjournals.org/content/70/9/3618.full#related-urlshttp://cancerres.aacrjournals.org/cgi/alertsmailto:[email protected]://cancerres.aacrjournals.org/content/70/9/3618http://cancerres.aacrjournals.org/

Cancer Research Selective Visualization of Cyclooxygenase-2 in … · Therapeutics, Targets, and...

Documents

Transcript of Cancer Research Selective Visualization of Cyclooxygenase-2 in … · Therapeutics, Targets, and...