Human Leukocyte Antigen Presented Macrophage Migration ......Companion Diagnostics and Cancer Biomarkers Human Leukocyte Antigen–Presented Macrophage Migration Inhibitory Factor

Companion Diagnostics and Cancer Biomarkers

Human Leukocyte Antigen–PresentedMacrophage Migration Inhibitory Factor Is aSurface Biomarker and Potential TherapeuticTarget for Ovarian CancerAndrea M. Patterson1, Saghar Kaabinejadian1, Curtis P. McMurtrey1,2,Wilfried Bardet1,Ken W. Jackson1, Rosemary E. Zuna3, Sanam Husain3, Gregory P. Adams4,Glen MacDonald4, Rachelle L. Dillon4, Harold Ames5, Rico Buchli2, Oriana E. Hawkins5,Jon A.Weidanz5, and William H. Hildebrand1,2

Abstract

T cells recognize cancer cells via HLA/peptide complexes, andwhen disease overtakes these immune mechanisms, immuno-therapy can exogenously target these same HLA/peptide surfacemarkers. We previously identified an HLA-A2–presented pep-tide derived from macrophage migration inhibitory factor(MIF) and generated antibody RL21A against this HLA-A2/MIFcomplex. The objective of the current study was to assess thepotential for targeting the HLA-A2/MIF complex in ovariancancer. First, MIF peptide FLSELTQQL was eluted from theHLA-A2 of the human cancerous ovarian cell lines SKOV3,A2780, OV90, and FHIOSE118hi and detected by mass spec-trometry. By flow cytometry, RL21A was shown to specificallystain these four cell lines in the context of HLA-A2. Next,partially matched HLA-A�02:01þ ovarian cancer (n ¼ 27) andnormal fallopian tube (n¼ 24) tissues were stained with RL21A

by immunohistochemistry to assess differential HLA-A2/MIFcomplex expression. Ovarian tumor tissues revealed significant-ly increased RL21A staining compared with normal fallopiantube epithelium (P < 0.0001), with minimal staining of normalstroma and blood vessels (P < 0.0001 and P < 0.001 comparedwith tumor cells) suggesting a therapeutic window. We thendemonstrated the anticancer activity of toxin-bound RL21A viathe dose-dependent killing of ovarian cancer cells. In summary,MIF-derived peptide FLSELTQQL is HLA-A2–presented andrecognized by RL21A on ovarian cancer cell lines and patienttumor tissues, and targeting of this HLA-A2/MIF complex withtoxin-bound RL21A can induce ovarian cancer cell death. Theseresults suggest that the HLA-A2/MIF complex should be furtherexplored as a cell-surface target for ovarian cancer immuno-therapy. Mol Cancer Ther; 15(2); 313–22. �2015 AACR.

IntroductionOvarian cancer is among the top five most deadly cancers in

women, with an estimated 14,180 deaths in the year 2015 for theUnited States alone (1). The majority of cases (61%) reach anadvanced stage before detection and, despite initial response tosurgery and platinum-based chemotherapy, have a 5-year survivalrate of only 27% (1). New approaches are pressingly needed, andtoday immunotherapeutic tumor targeting strategies such as

monoclonal antibody therapy or adoptive transfer of cancer-directed T cells are proving increasingly effective (2–4). Forexample, the anti-HER2 antibody trastuzumab is significantlyextending survival in breast cancer patients (5), and CD19-direct-ed chimeric antigen receptor (CAR) T-cell therapies are reaching90% complete response rates in multiple clinical trials of B-cellacute lymphoblastic leukemia (6). These exciting breakthroughsindicate the value of immune therapies, but truly successfulextension of these strategies to several cancer types, includingovarian, thus far remains elusive.

One element common to these and other immuno-oncologysuccesses is the exploitationof effective cancer cell-surfacemarkersfor direct tumor targeting. A handful of candidate markers havebeen explored clinically in ovarian cancer, such as epithelial celladhesion molecule (EpCAM) or EGFR as antibody targets, orfolate receptor-a (FRa) as a target for CAR T-cell therapy (7, 8).Early trials with these targets are proving somewhat efficacious,especially in combination with chemotherapies, and as immu-notherapies become more refined, these targets may proveincreasingly effective. An important consideration in targetedtherapy is the significant heterogeneity that is particularly appar-ent in ovarian cancer (9). Even when patient-specific, a singleantigen will rarely be sufficient for eradicating ovarian tumors;successful strategies will likely target multiple patient-relevant

1Department of Microbiology and Immunology, University of Okla-homa Health Sciences Center, Oklahoma City, Oklahoma. 2Pure MHCLLC,Oklahoma City,Oklahoma. 3Department of Pathology, Universityof Oklahoma Health Sciences Center, Oklahoma City, Oklahoma.4Viventia Bio Inc., Winnipeg, Manitoba, Canada. 5Department ofImmunotherapeutics andBiotechnology,TexasTechUniversity HealthSciences Center, Abilene, Texas.

Note: Supplementary data for this article are available at Molecular CancerTherapeutics Online (http://mct.aacrjournals.org/).

Corresponding Author: William H. Hildebrand, University of Oklahoma HealthSciences Center, 975 NE 10th Street, BRC-317, Oklahoma City, OK 73104. Phone:405-271-1203; Fax: 405-271-3117; E-mail: [email protected]

doi: 10.1158/1535-7163.MCT-15-0658

�2015 American Association for Cancer Research.

MolecularCancerTherapeutics

www.aacrjournals.org 313

on June 28, 2021. © 2016 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst December 30, 2015; DOI: 10.1158/1535-7163.MCT-15-0658

http://mct.aacrjournals.org/

markers at once in an effort to prevent the escape mutants thatspawn recurrence. As immunotherapy strategies continue toevolve, new cell-surface antigens are needed toward effectiveovarian cancer targeting.

Numerous molecular changes distinguish cancerous cells fromhealthy cells, but the vast majority of these are hidden inside theplasma membrane and inaccessible to antibody or CAR T-celltherapies. As such, HLA molecules play a pivotal role exposingtumor cells for immune recognition. In ovarian cancer, dysregu-lated intracellular proteins are naturally targeted at the cell surfacewhen T cells recognize class I HLA molecules presenting altered-self peptides; the presence of tumor-infiltrating CD8þ lympho-cytes corresponds with significantly longer survival for ovariancancer patients, demonstrating this phenomenon (10–12). This isthe basis for cancer vaccine strategies using peptides from tumorantigens, such as NY-ESO-1, HER2, p53, or MUC-1, as well asimmune checkpoint inhibitors that unleash T cells for recognitionof tumor-associatedHLA/peptide complexes (7, 8). In addition tovaccine elicited, chimeric, and checkpoint-released T cells, a newclass of monoclonal antibodies that mimic the T-cell receptor areable to single outHLA/peptide complexes thatmark the surface ofcancerous cells (13, 14). In this way, HLA/peptide complexesreveal a wealth of intracellular protein changes that can beaccessed by monoclonal antibodies and derivative therapeuticstrategies.

Macrophage migration inhibitory factor (MIF) is a proinflam-matory cytokine with dramatic overexpression and tumor-pro-moting properties in ovarian cancer (15–18). MIF production byovarian tumors correlates with increased macrophage infiltrationand stimulatesmacrophage functions that in turn promote tumorinvasiveness (15, 17). MIF also appears to support ovarian tumorsurvival and proliferation through p53 inhibition and Akt acti-vation, and to promote angiogenesis through TNFa, IL6, andVEGF (17, 19). Furthermore, increased MIF expression in bothclassical ovarian cancer cells and in ovarian cancer stem cellssuggests it may be developed as a marker for both (20, 21).Strategies targeted to MIF in various contexts have includedsmall-molecule inhibitors, neutralizing antibodies, and siRNA(22–24). However, the discovery of a MIF-derived peptide that ispresented by the class I HLA of cancer cells offers the uniqueopportunity to target MIF in the form of a cancer cell-surfaceantigen (25, 26). The hypothesis tested here is that ovarian cancerwill be susceptible to targeting through the HLA-presented MIFpeptide. To test this hypothesis, we first demonstrated proteomi-cally that MIF peptide FLSELTQQL can be consistently eluted andsequenced from the HLA of ovarian cancer cell lines. The subse-quent proficiency of the HLA/MIF-specific antibody RL21A inselectively staining ovarian cancer tissue and in mediating toxin-conjugated killing of ovarian cancer cells further implicates HLA/MIF as a cell-surface biomarker and potential target for exploita-tion in ovarian cancer immunotherapy.

Materials and MethodsCell lines

SKOV3 cells were purchased from the ATCC in 2011 and grownin DMEM/F12K medium with 10% FBS. A2780 cells were pur-chased from Sigma in 2012 and grown in RPMI with 10% FBS.FHIOSE-118hi cells andOV90 cells were received in 2011 as kindgifts from Dr. Patricia Kruk, University of South Florida (Tampa,FL), and were both grown in 1:1 Medium 199/MCDB 105 with

15% FBS. FHIOSE-118hi is an artificially transformed humanovarian cell line that is tumorigenic inmice (27), originating fromthe lab of Dr. Kruk and hereafter referred to as FHIOSE, andOV90cells were purchased from the ATCC in 2001 and sent to our labfrom a low-passage frozen stock. T2 cells generated by Dr. PeterCresswell, Duke University (28), were received as a kind gift fromthe Cresswell lab and grown in RPMI with 10% FBS. All cell lineswere subjected to high-resolution sequence-based HLA typing(HLA-A, -B, -C, and -DRB1) immediately upon receipt and growthin our laboratory, and then again after stable transfectionand subcloning to ensure authentication prior to use in datacollection.

Production of soluble HLA-A�02:01 and peptide purificationSoluble HLA (sHLA) was produced and peptides were purified

as previously described (29). Briefly, each cell line was stablytransfected with the sHLA-A�02:01 construct, subcloned, andgrown at high density in hollow-fiber bioreactors. The sHLA/peptide complexes were isolated from supernatants over animmunoaffinity column using an engineered C-terminal purifi-cation tag. Peptides were removed by acid boil, purified through a3-kDa filter, and lyophilized.

LC/MSFirst dimension LC. Peptides were resuspended in 10% acetic acidand fractionated over a 150-mm longC18 column (2-mm internaldiameter Gemini column; pore size, 110 Å; particle size, 5 mm;Phenomenex) for reverse-phase high-performance LC (RP-HPLC)using a Michrom BioResources Paradigm MG4 instrument. Anacetonitrile (ACN) gradient was run at pH 10 using a two-solventsystem. Solvent A contained 2% ACN in water, and solvent Bcontained 95% ACN in water, each with 10 mmol/L ammoniumformate. The column was pre-equilibrated at 2% solvent B.Peptide was loaded at a flow rate of 120 mL/min over an 18-minute period, and then two linear gradients were run at160 mL/min: 4% to 40% B for 40 minutes, followed by 40%to 80% B for 8 minutes.

Second dimension LC. Forty peptide-containing first dimensionHPLC fractions were dried, resuspended in 10% acetic acid,and applied individually to nanoscale RP-HPLC (EksigentnanoLC415; AB Sciex), including a 0.5-mm long C18 trapcolumn (350-mm internal diameter; ChromXP) with 3-mmparticles and 120Å pores, and a 15-cm-long C18 separationcolumn (75-mm internal diameter; ChromXP) packed with thesame medium. An ACN gradient was run at pH 2.5 using atwo-solvent system. Solvent A was 0.1% formic acid in water,and solvent B was 0.1% formic acid in 95% ACN in water. Thecolumn was pre-equilibrated at 2% solvent B. Samples wereloaded at 5 mL/min flow rate onto the trap column and runthrough the separation column at 300 nl/min with two lineargradients: 10% to 40% B for 70 minutes, followed by 40% to80% B for 7 minutes.

MS. The column effluent was injected using the nanospray III ionsource of an AB Sciex TripleTOF 5600 quadrupole time-of-flightmass spectrometer with the source voltage set to 2,400 v. Infor-mation-dependent analysis (IDA) of peptide ions was acquiredbased on a survey scan in the TOF-MS positive-ion mode over arange of 300 to 1,250m/z for 0.25 seconds. Following each survey

Patterson et al.

Mol Cancer Ther; 15(2) February 2016 Molecular Cancer Therapeutics314




scan, up to 22 ions with a charge state of 2 to 5 and intensity of atleast 200 counts were subjected to collision-induced dissociation(CID) for tandemMS analysis (MS/MS) over a maximum periodof 3.3 seconds. Selection of a particular ion m/z was excluded for30 seconds after three initial MS/MS experiments. Dynamiccollision energy was utilized to automatically adjust the collisionvoltage based upon ion size and charge. PeakView Softwareversion 1.2.0.3 was used for data visualization.

AntibodiesThe BB7.2 (pan-HLA-A2 monoclonal antibody) expressing

mouse hybridoma cell line was purchased from the ATCC. TheRL21A hybridoma was generated by PureMHC as previouslydescribed (26). BB7.2 and RL21A hybridomas were grown inHybridoma-SFM (Invitrogen), and antibodies were purified usingproteinG(GEHealthcare).Mouse IgG2a isotype control antibodywas purchased fromR&D Systems (20102). Mouse IgG2b isotypecontrol antibody was purchased from BioLegend (MG2b-57).Goat anti-mouse IgG AffiniPure F(ab0)2 fragment conjugated toR-Phycoerythrin was purchased from Jackson ImmunoResearchLaboratories, Inc. (115-116-071).

Flow cytometrySKOV3 and A2780 cells were first transfected with a full-length

HLA-A�02:01 construct as previously described (30), and theresulting SKOV3-A2 and A2780-A2 along with SKOV3, A2780,FHIOSE, andOV90 cells (5� 105) were stained with 0.01mg/mLof RL21A, BB7.2, or isotype controls for 30 minutes followed byphycoerythrin (PE)-conjugated goat anti-mouse IgG for 30 min-utes. Cells were analyzed on a Stratedigm S1200Ex flow cyt-ometer, and data analysis was performed using FlowJo Software(Tree Star).

RL21A peptide specificity T2 assayTAP-deficient T2 cells expressing HLA-A�02:01 (1 � 106) were

pulsed in AIM-V Medium (Gibco) for 4 hours with relevantFLSELTQQL peptide or irrelevant HLA-A2–binding VLQGVLPAL,YLEPGPVTV, YLLEMLWRL, AIMDKNILL, GVLPALPQV, KIGEG-TYGV, TMTRLLQGV, ALMPVLNQV, YELPGPVTV, SVGGVFTSV,and NLVPMVATV peptides at 20 mmol/L. The pulsed T2-A2 cellswere stained with 1 mg/mL RL21A, BB7.2, or isotype control,followed by the PE-conjugated goat anti-mouse IgG. Cells wereanalyzed on FACSCanto II (BD Bioscience). Data analysis wasperformed using FlowJo Software (Tree Star).

HLA cross-reactivity screenAssay generation. Production and purification of 120 differentrecombinant intact class I sHLA molecules were performed asdescribed previously with minor modifications (31). To immo-bilize the purified sHLA molecules onto a solid Luminex beadsupport, they were first tagged with the biotin derivative NHS-PEG12-Biotin (Thermo Fisher Scientific) according to the man-ufacturer's instructions. An empirically determinedmolar ratio of20:1 (Biotin-NHS:sHLA protein) was applied so that on averagenot more than one tag was bound to each sHLA protein. In a totalvolume of 10 mL, a titrated concentration of individual sHLAprobes (to give 95% loading efficiency) wasmixed with 2.5� 106LumAvidin beads (L100-Lxxx-01; Luminex) in PBSþ 1%BSA andincubated for 60 minutes with light shaking. Beads were washed4� in PBS (spins 10minutes at 5,300� g) and resuspended in PBS

þ 1% BSA þ 0.05% sodium azide for storage. To create themultiplex assay, beads were combined into two sets: A/C allelesand B alleles. A panel of 25 controls was also created, includinghuman immunoglobulins (IgG, IgM, IgA, IgE, and IgG1/2/3/4),mouse immunoglobulins (IgG1, IgG2a, ND IgM), kappa, lamb-da, b2-microglobulin, BGG, BSA, HSA, as well as HLA-G and E,CD1c, and C1q.

RL21A screen. The target peptide for RL21A (FLSELTQQL)was incubated with the sHLA-Luminex beads in 6,000-foldexcess at 53�C for 15 minutes, followed by 48 hours at 4�Cfor complex reassembly with peptide. (Note that sHLA withincompatible binding motifs for this peptide will not replacethe original endogenous peptides.) RL21A (25 mg/mL) wasincubated with each set of beads for 60 minutes at 4�C,washed four times and then labeled with anti-mouse Fcconjugated to PE, and incubated for a further 60 minutesat 4�C. Fluorescence intensity and bead labels were visualizedon a Luminex 100 reader. Data were collected for a minimumof 400 antigen-coated beads. Reported values represent medi-an fluorescence intensity after subtraction of the secondary-only background signal determined for each bead individu-ally and adjusted by applying the internal negative controlbeads as a lower threshold value.

Human tissuesCryopreserved human tissue samples were acquired from the

University of Oklahoma Stephenson Cancer Center BiospecimenCore (OklahomaCity, OK) following approval from the ScientificReview Committee (#200908-007) and the Institutional ReviewBoard (#3159). Samples were chosen based on a diagnosis ofserous ovarian carcinoma, as this is the most common subtype,and HLA-A�02:01þ tissues were selected by high-resolutionsequence-based HLA typing of matched blood samples. A totalof 84 blood samples were HLA typed, and 39 HLA-A�02:01þtissues or tissues pairs (tumor and/or fallopian tube, FT) wereobtained. Hematoxylin and eosin (H&E) staining was used toverify tissue identities and the presence of evaluable cells. Asindicated in Table 1, 27 serous ovarian carcinoma tissues and24 normal FT tissues were evaluable, 14 pairs of which werematched tumor and normal. Patient characteristics are summa-rized in Table 1.

ImmunohistochemistryCryopreserved tissues were sectioned at 6 mm, fixed with 5%

methanol in PBS for 12minutes, and then allowed to air dry for atleast 15 minutes. Tissues were encircled with a PAP pen andblocked for 20 minutes with normal horse serum (Vector Labs).Primary antibodies were added for 45 minutes: BB7.2 (pan-HLA-A2) at 1 mg/mL, RL21A (HLA-A2/MIF) at 2 mg/mL, or mouseIgG2a isotype control at 2 mg/mL. An horseradish peroxidase–conjugated anti-mouse Ig secondary was applied for 30 minutes(ImmPRESS reagent; Vector Labs). Slides were developed for 4minutes with DAB, counterstained with hematoxylin, and dehy-drated through changes of 80%, 95%, and 100% reagent alcohol,then xylene, and coverslipped with nonaqueous mounting medi-um. Tissues were scored on a sliding scale from 0 to 3 for stainingintensity and from0 to4 for percentage of cells stained (0, none; 1,1%–10%; 2, 11%–30%; 3, 31%–60%; 4, 61%–100%). These twovalues were multiplied together to give a composite score of 0 to

The HLA/MIF Complex as a Surface Target for Ovarian Cancer

www.aacrjournals.org Mol Cancer Ther; 15(2) February 2016 315




12. The isotype control was used to subtract background staining.Two gynecologic pathologists (R.E. Zuna and S. Husain) scoredthe tissues independently, and scores were averaged. Scoring wasperformed from the original slides using OLYMPUS BX43 micro-scopes, and images were taken with a NIKON NI-U and NIS-Elements software (Nikon).

Cytotoxicity assaysWhere unconjugated antibodywas used, cells were treatedwith

various concentrations of RL21A (0.1 pmol/L–10 nmol/L) alongwith 20 nmol/L cytotoxic secondary antibody (Fab fragment ofanti-mouse IgG Fc conjugated with Duocarmycin DMwith cleav-able linker; Moradec LLC) for 72 hours. Alternatively, RL21A wasdirectly conjugated to the deantigenized plant toxin deBouganin(32, 33) using amethod described previously (34), and cells weretreated with various doses of the immunotoxin (0.008 pmol/L–80 nmol/L) for 72 hours. MTS reagent was then added at 20 mL/well and incubated for 3hours.Optical density (OD)was detectedat 490 nm by a SpectraMax 190 Microplate Spectrophotometerand viewed using SoftMax Pro 5 Software (Molecular Devices).The percentage of live cells was calculated by dividing the OD ofeach well by the average OD of control wells untreated withprimary antibody (but treated with secondary toxin) or untreatedwith conjugated antibody.

Statistical analysisGraphing and statistical analysis were performed using Graph-

Pad Prism 6 Software version 6.0e.

Results and DiscussionMIF19–27 is presented by the HLA-A2 of ovarian cancer cell lines

The MIF-derived peptide FLSELTQQL (MIF19–27) was firstidentified as a breast cancer biomarker in the context of thecommon allotype HLA-A�02:01 (25). Here, we investigatedwhether the same peptide might also mark ovarian cancer.Because the processing and subsequent HLA presentation ofantigens rely upon a multifactorial system that is not fully under-stood, a priori one cannot link protein expression with peptidepresentation (35, 36). Thus, despite the established overexpres-sion ofMIF protein in ovarian cancer, it is difficult to knowwhich,if any, MIF peptides might be presented. Therefore, we harvestedHLA-A�02:01 at high yields from the cancerous human ovariancell lines A2780, SKOV3, OV90, and FHIOSE to proteomicallyassess MIF19–27 peptide presentation. Eluted peptides from eachof these lines were compared with a synthetic MIF19–27 peptidethat acted as a reference for LC elution time andMS fragmentationpattern (Fig. 1A and B). For each cell line, an eluted peptide peakof the appropriate mass (m/z 539.79), elution time (40 � 2minutes), and fragmentation pattern unequivocally confirmedthe natural processing and presentation of the MIF19–27 peptide(Fig. 1A andC). These data demonstrate that theoverexpressionofthe MIF protein across ovarian cancer cell lines and tissues in theliterature coincides with consistent HLA-A2 presentation ofMIF19–27 among four different cancerous ovarian cell lines.

RL21A recognizes HLA-A2þ ovarian cancer cellsHaving confirmed that MIF19–27 is naturally prepared for HLA-

A2presentation inovarian cancer cells, we proceeded to test surfacestaining of the same cell lines with the monoclonal antibodyRL21A. RL21A is specific to the HLA-A2/MIF19–27 complex anddoes not bind HLA-A2 presenting various irrelevant peptides(ref. 26 and Supplementary Fig. S1). First, FHIOSE cells thatnaturally express HLA-A�02:01 were stained with the pan-HLA-A2 antibody BB7.2 confirming robust HLA-A2 surface expression(Fig. 2A and F). These cells then also demonstrated substantialRL21A staining, suggesting that considerable fractions of theirHLA-A2 molecules are loaded with the MIF19–27 peptide and that HLA-A2/MIF19–27 is a naturally occurring and high abundance surfacebiomarker on these cells (Fig. 2B and F). As A2780 and SKOV3 donot naturally express HLA-A2, these cell lines were transfected toexpress a full-lengthHLA-A�02:01 surfacemolecule (Fig. 2A). TheseA2780-A2 and SKOV3-A2 cells also stained efficiently with RL21A(Fig. 2C,D, and F). Thus, various ovarian cancer cell lines all appearto make MIF19–27 available in HLA-A2 at the cell surface.

RL21A antigen cross-recognition benefits population coverage.Receptors of the adaptive immune system are empowered by finespecificity, yet both monoclonal antibodies and T-cell receptorscan at times be promiscuous in their recognition of antigen. Anexample is seen in Fig. 2F where the pan-HLA-A2 antibodyexhibits a known cross-reactivity with a close serologic relative,A�68:01, which is endogenous to the untransfected SKOV3 cells.In therapeutic development, immune cross-recognition of similarantigensmust be investigated to avoid off-target effects, and this isaddressed in the next section. However, limited cross-reactivitymay be beneficial if the same cancer-associated peptide can berecognized in the context of multiple HLA molecules, wideningthe relevant patient population. The OV90 cells that do notexpress HLA-A�02:01 are a case in point, as they encode the

Table 1. Donor tissue characteristics

n (%)

Total blood samples HLA-typed 84Diagnosis: serous ovarian carcinoma 84/84 (100)HLA-A�02:01 allotype 39/84 (46)

Total HLA-A�02:01 tumor tissues 27FIGO pathologic stageI 3/27 (11)II 2/27 (7)III 20/27 (74)IV 0/27 (0)Not determined 2/27 (7)

Two-tier gradeLow 2/27 (7)High 23/27 (85)Not determined 2/27 (7)

Tumor tissue regions evaluableEpithelial carcinoma cells 27/27 (100)Stroma 26/27 (96)Blood vessel endothelium 24/27 (89)

Total HLA-A�02:01 normal FT tissues 24Normal tissue regions evaluableEpithelium 19/24 (79)Stroma 24/24 (100)Blood vessel endothelium 24/24 (100)

Of above, matched tumor and FT tissue pairs 14Matched tissue regions evaluableEpithelium/epithelial carcinoma cells 11/14 (79)Stroma 14/14 (100)Blood vessel endothelium 12/14 (86)

NOTE: Gray boxed, corresponding to tissue numbers in Fig. 4.Abbreviation: FIGO, International Federation of Gynecology and Obstetrics.

Patterson et al.





closely relatedHLA-A�02:02 and A�02:05 alleles and demonstrat-ed cross-reactivity with both BB7.2 and RL21A (Fig. 2A, E, and F).This OV90 reactivity suggests that MIF19–27 is able to bind one orboth of the alternate HLA-A2 alleles for recognition by RL21A. Tomore fully characterize which HLA allotypes might be relevant toMIF19–27 and RL21A, microspheres were coated with a panel of120 class I HLA molecules and subjected to MIF19–27 peptideloading and staining with RL21A. A degree of cross-reactivity wasobserved among all of the HLA-A�02:0x isoforms tested with noreactivity toward any of the other HLA-A, -B, or -C molecules(Supplementary Fig. S2 and data not shown). In the event thatRL21A demonstrates promise as a cancer immunotherapy, rec-ognition of MIF19–27 in the context of multiple HLA-A2 subtypesprovides wider HLA class I population coverage.

RL21A selectively stains ovarian cancer tissuesOvarian cancer cell lines represent a valuable model of disease,

and the panel of cell lines tested here demonstrated consistentHLA-A2 presentation ofMIF19–27 by proteomic analysis as well asby RL21A surface staining. Next, in order to test the naturaltumor environment while at the same time testing for off-target

reactivity, RL21A was used to directly stain ex vivo normal andtumor tissue samples from patients with serous ovarian cancer(Table 1). FT epithelium is increasingly considered the primarysite from which the majority of serous ovarian cancers develop(37–39), and thus this cell type was evaluated as the primarynormal comparison for RL21A staining. The comparisonrevealed dramatically increased RL21A staining intensities inthe ovarian cancer cells with an intensity score � 1þ observedin 67% (18/27) of tumors compared with 5% (1/19) of thenormal epithelial tissues, and a score of �2þ was seen in 41%(11/27) of tumor tissues compared with none of the normalepithelia (Fig. 3A and B). Incorporating intensity and percent-age of cells stained into overall staining scores for each tissue, asignificant increase in RL21A reactivity was found in ovariantumors compared with normal epithelial tissues (median, 0.25vs. 3.50; P < 0.0001; Fig. 4A). In patients where both tumor andnormal FT epithelial cells were evaluable (see Table 1), anupward slope between matched data points demonstrates anincrease in staining upon cancerous transformation (Fig. 4B).The relatively high percentage of tumor tissues positive for thiscomplex corresponds to the consistent MIF19–27 sequencing

Figure 1.The MIF peptide FLSELTQQL is detected in HLA-A2 from all four cancerous ovarian cell lines. A, overlaid extracted ion chromatograms of the 539.79m/z ion fromthe synthetic MIF19–27 LC-MS run and that of the corresponding fraction from each cell line. B, MS/MS spectrum showing the fragmentation pattern of thesynthetic MIF19–27 peptide. B and Y ions are labeled along with some prominent internal fragment ions. C, the identical fragmentation pattern to the synthesizedMIF19–27 peptide in B was observed in all ovarian cancer cell lines at the corresponding m/z.






and RL21A staining of cell lines (Figs. 1 and 2) as well as theprevalent overexpression of the MIF protein observed acrossovarian tumors (17, 18). This dramatic transformation-associ-ated increase in RL21A epithelial staining demonstrates that theHLA-A2/MIF19–27 complex represents an ovarian tumor–spe-cific biomarker on patient specimens.

Prevalent HLA upregulation was observed in ovarian tumors. HLAdownregulation has been reported as a means of immune escapein several cancers, but the extent to which this occurs is contro-versial. As a control along with RL21A staining, tissue sections inthis study were stained with BB7.2 to confirm expression of HLA-A2 (Fig. 3A; Supplementary Fig. S3). Unexpectedly, an overallincrease in total HLA-A2 expression was detected in the tumorcells compared with the normal FT epithelium (median, 5.00 vs.7.50; P < 0.05; Fig. 4A). Ten of the 11 matched cases displayed anincrease in HLA-A2 staining score between normal and cancerousepithelial cells, whereas only one displayed HLA-A2 downregula-tion (Fig. 4B). The literature for ovarian cancer reports HLAdownregulation in as few as �20% of cases and in as many as�70% (10, 40–43). This variation could be attributed to differingantibodies, tissue preservation and fixation methods, cohortcharacteristics, and various technical definitions of "downregula-

tion." In comparison, this is the first report to our knowledgewhere ovarian cancer HLA class I expression was studied usingcryopreserved tissues. A delicate fixation buffer (5% methanol)was used to avoid HLA denaturation as BB7.2 and RL21A anti-bodies require intact HLA/peptide complexes, an approach thatcontrastswith traditional immunohistochemistry (IHC)methodsusing strongly fixed, paraffin-embedded tissues and antibodiesthat recognize only a denatured HLA conformation (44). Thestrong correlation between cytotoxic T-cell tumor infiltration andsurvival in ovarian cancer also suggests that the class I HLA systemis largely operative and that immune escape hinges predominant-ly on effector cell inhibition rather than on HLA downregulation(10–12). Our observations suggest a low incidence of HLA down-regulation and reveal an increase in HLA-A2 expression in >90%of ovarian tumor tissues (Fig. 4B). These findings enhance theprospect of therapies for ovarian cancer that directly target HLA/peptide complexes on the tumor, including monoclonal antibo-dies or CAR T cells directed against HLA-A2/MIF19–27.

RL21A recognizes the tumor stroma. In addition to the epithelialcomparison, stromal tissue and blood vessel endothelial cells wereseparately evaluated for RL21A and BB7.2 reactivities. RL21A stain-ing increased significantly in the tumor stroma (median, 0.50 vs.

Figure 2.RL21A stains ovarian cancer cells in an HLA-A2–dependent manner. A, overlaid histograms comparing total HLA-A2 staining for each cell line. B, flowcytometry dot plots showing RL21A staining of FHIOSE compared with isotype control. C and D, RL21A staining of HLA-A2–transfected A2780-A2 (C, left) orSKOV3-A2 (D, left) compared with the untransfected parent cells (right). E, RL21A staining of OV90 compared with isotype control. F, summarized RL21Aand pan-HLA-A2 mean fluorescence intensities for all lines from three independent experiments. Data are mean þ SD.

Patterson et al.





2.75; P < 0.0001), and among the different cell types, the tumorstromaexhibited thebiggest difference in totalHLA-A2expressionascompared with its normal counterpart (median, 2.50 vs. 5.88;P < 0.001; Fig. 4C). Tumor stroma is increasingly recognized as aprominent player in ovarian cancer progression, and multipletherapeutic strategies are now targeting this compartment (45).Furthermore, increased MIF expression is reported in the ovariantumor stroma as contributing to tumor promotion (46). A thera-peutic that can target both tumor epithelium and tumor stromawouldbeof great value, and the IHCdatapresentedhere suggest thatboth cell types process and present the MIF19–27 peptide for recog-nitionbyRL21A. In contrast, bloodvessel endotheliumyieldedhighHLA-A2, but low RL21A, staining in both normal and tumor tissues(Fig. 4D). Thus,RL21Acanefficiently andspecifically reactwithbothovarian tumor cells and the associated cancer stroma.

Potential for a therapeutic window. As in the example of HER2,successful targeting of cell-surface tumor antigens requires asignificant increase in antigen availability on the cancerous cellsin comparison with normal tissues. Just as tumor availability ofHLA-A2/MIF19–27 dramatically increases in comparison with nor-mal epithelium (Fig. 4A), it is also significantly higher in com-parison with the stromal and endothelial tissues of the normal FTsamples (P < 0.001 and P < 0.0001, respectively; Fig. 4E). Thus, allof the normal tissue regions observed here remained significantlybelow tumor RL21A staining levels. These findings correspondwith previous RL21A cadaveric studies inwhich no significant off-target staining was observed across multiple normal human

organs (26). In the same study, fresh peripheral blood sampleswere also tested from healthy volunteers with little to no RL21Astaining observed across 18 donors (26). A clear visual represen-tation of RL21A tumor specificity is seen in Fig. 4F, an example ofwhenmetastatic lesionswere found embeddedwithin the normalFT tissue of ovarian cancer patients. RL21A strongly reacts with thetumor focus, yet shows no reactivity with directly adjacent normalepithelial cells. These data demonstrate that a significant thera-peutic window may exist for RL21A targeting of ovarian cancer.

Toxin-conjugated RL21A mediates killing of ovarian cancercells

Toxin conjugation is a promising avenue for converting tumor-specific antibodies into potent anticancer therapeutics (47). Forinstance, a toxin-conjugated version of trastuzumab significantlyimproves progression-free survival in breast cancer trials, includinga decrease in adverse events (48). To begin assessing therapeuticpotential of RL21A, A2780-A2 cells were treated with increasingRL21A concentrations alongside the HLA-A2–negative A2780parent cells in the presence of a constant concentration of second-ary Fc-binding duocarmycin-based cytotoxin. Toxin-associatedRL21A killed the A2780-A2 cells in a concentration-dependentmanner with an IC50 of 0.048 nmol/L (R

2 ¼ 0.945) while cellslackingHLA-A2wereunaffected (Fig. 5A).Next, RL21Awasdirectlyconjugated to deBouganin, a deimmunized plant toxin derivedfrom Bougainvillea spectabilis (32, 33). This conjugate (RL21A-deB)exhibited dose-dependent cytotoxicity against A2780-A2 with anIC50 of 0.324 nmol/L (R

2 ¼ 0.970; Fig. 5B). To also test

Figure 3.RL21A tissue staining intensities. A,four different stains were performedon each patient tissue as indicatedalong the top. Three tissues are shown,each representing the indicated RL21Aintensity range where x ¼ intensity.Original magnification, �200(including zoomed insets). B,percentages of tumor and normalepithelial tissues falling within eachRL21A intensity range. Actual numberof tissues is indicated above each bar.






effectiveness against cells that naturally express HLA-A�02:01,FHIOSE cells were exposed to the same concentrations ofRL21A-deB, anda similar cytotoxicitypatternwasobservedwithanIC50 of 9.376 nmol/L (R

2 ¼ 0.925; Fig. 5B). These data correlatewellwith the relative levels ofRL21A stainingobserved for each celllinebyflowcytometry (Fig. 2F). Thus,HLA-A2presentsMIF19–27 ata level whereby toxin-bound RL21A kills cancerous ovarian cellswithout harming cells that lack the antigen. RL21A-deB immuno-toxin represents one of the potential avenues toward therapeutictargeting of HLA-A2/MIF19–27 complexes on ovarian cancer.

SummaryImmune therapies for cancer are enjoying a renaissance.Mono-

clonal antibodies continue to evolve into potent anticancer ther-

apies via drug conjugation, through bi-specific technologies, andusing antibody-variable regions to engineer highly active CAR Tcells against cell-surface tumor markers. In addition, HLA-pre-sented cancer vaccines are stimulating endogenous T cells againsttumor antigens, whereas immune checkpoint inhibitors can fur-ther unleash bothmanipulated and natural effector T cells (7). Asimmunologists and cancer biologists join forces to harness thespecificity and power of immunotherapies, the need for a robustsource of cell-surface cancer markers becomes increasingly appar-ent; T cells and antibodies are only as powerful as their targets.Cell-surfaceHLA class Imolecules are a naturalmeans of revealingintracellular indicators of neoplasticity to the immune system,and given the observation that tumor-infiltrating CD8þ lympho-cytes correspond to better clinical outcomes, we have embarkedupon targeting ovarian cancer via class I HLA-presented peptides.

Figure 4.RL21A specifically targets ovarian tumor tissues. Normal FT or tumor tissues from ovarian cancer patients were stained by IHC for total HLA-A2 using BB7.2 orfor the specific HLA-A2/MIF complex using RL21A. Scores integrate intensity and percentage of cells stained. Individual data points and median are plottedfor total epithelial scores (A), matched-only epithelial scores with normal and tumor pairs connected (B), stromal scores (C), and blood vessel endothelialscores (D). Mann–Whitney U tests were performed for all but B, which represents a Wilcoxon matched-pairs signed rank test. E, RL21A scores only, comparingnormal FT stromal and endothelial cells with the epithelial ovarian cancer cells using a Kruskal–Wallis one-way ANOVAwith Dunn's multiple comparisons. F, normalFT tissue with tumor focus boxed and further magnified. Black arrows, epithelial ovarian cancer metastasis. White arrows, normal epithelium. Originalmagnification, �40 (top), �200 (bottom). � , P < 0.05; ��, P < 0.01; ��, P < 0.001; �� , P < 0.0001.

Patterson et al.





Here, we explored the availability of the MIF-derived peptideFLSELTQQL in the context of ovarian cancer HLA-A�02:01 mole-cules. We found that tumorigenic ovarian cell lines expressingHLA-A2display this preciseMIF19–27 peptide and are reactivewiththeRL21Aantibody specific for this complex.When cryopreservedtumor samples from ovarian cancer patients were tested forreactivity, strong staining for HLA-A2/MIF19–27 complexes wasobserved, which was significantly increased from normal FTepithelium, stroma, and endothelium. An unexpected increasein HLA-A2 expression was noted in the tumors, and alternateHLA-A2 variants were recognized as capable of presenting theMIFpeptide and reacting with this antibody. As monoclonal reagents

are enabling a number of immunotherapies for cancer, we toxin-conjugated RL21A and demonstrated its potential for specifickilling of ovarian cancer lines in vitro. These data merit the futuretesting of a humanized version of RL21A to confirm continuedcytotoxicity in vitro, to investigate efficacy in vivo using animalmodels of human ovarian cancer and, if results continue to bepromising, towork toward a transition to clinical trials for ovariancancer immunotherapy.

Disclosure of Potential Conflicts of InterestW.H. Hildebrand is Chief Scientist at and has ownership interest (including

patents) in Pure Protein. No potential conflicts of interest were disclosed by theother authors.

Authors' ContributionsConception and design: A.M. Patterson, S. Kaabinejadian, R. Buchli,O.E. Hawkins, W.H. HildebrandDevelopment of methodology: A.M. Patterson, S. Kaabinejadian, C.P. McMur-trey, R.E. Zuna, G. MacDonald, R.L. Dillon, R. Buchli, O.E. Hawkins,J.A. Weidanz, W.H. HildebrandAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A.M. Patterson, S. Kaabinejadian, C.P. McMurtrey,K.W. Jackson, R.E. Zuna, R.L. Dillon, H. Ames, R. Buchli, J.A. Weidanz,W.H. HildebrandAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): A.M. Patterson, C.P. McMurtrey, S. Husain, H. Ames,R. Buchli, J.A. WeidanzWriting, review, and/or revision of the manuscript: A.M. Patterson, K.W.Jackson, R.E. Zuna, S. Husain, G.P. Adams, G. MacDonald, H. Ames, R. Buchli,W.H. HildebrandAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): A.M. Patterson, W. Bardet, K.W. Jackson,R. Buchli, J.A. WeidanzStudy supervision: W.H. Hildebrand

AcknowledgmentsThe authors thank the Stephenson Cancer Center at the University of

Oklahoma Health Sciences Center, Oklahoma City, OK, for the use of boththe Biospecimen Acquisition Core and Bank, which provided the cryopreservedtissues, and the Cancer Tissue Pathology Core, which provided cryosectioningand H&E staining services (special thanks to Dr. David Brown and LouisaWilliams, respectively).

Grant SupportThis study was supported in part by an NIHNRSA NIAID Ruth L. Kirschstein

Training Grant T32AI007633 "Molecular Basis of Immunity" (to A.M. Patter-son) and by Emergent Technologies Inc.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received August 11, 2015; revised October 9, 2015; accepted November 7,2015; published OnlineFirst December 30, 2015.

References1. American Cancer Society. Cancer facts and figures 2015. Atlanta,

2015.2. Disis ML. Mechanism of action of immunotherapy. Semin Oncol 2014;41

Suppl 5:S3–13.3. Trivedi S, Jie HB, Ferris RL. Tumor antigen-specific monoclonal antibodies

and induction of T-cell immunity. Semin Oncol 2014;41:678–84.4. JooWD, Visintin I,Mor G. Targeted cancer therapy–are the days of systemic

chemotherapy numbered? Maturitas 2013;76:308–14.5. Hudis CA. Trastuzumab–mechanism of action and use in clinical practice.

N Engl J Med 2007;357:39–51.

6. Bouhassira DC, Thompson JJ, Davila ML. Using gene therapy to manip-ulate the immune system in the fight against B-cell leukemias. Expert OpinBiol Ther 2015;15:403–16.

7. Chester C, DorigoO, Berek JS, Kohrt H. Immunotherapeutic approaches toovarian cancer treatment. J Immunother Cancer 2015;3:7.

8. Wefers C, Lambert LJ, Torensma R, Hato SV. Cellular immunotherapy inovarian cancer: targeting the stem of recurrence. Gynecol Oncol2015;137:335–42.

9. Meinhold-Heerlein I, Hauptmann S. The heterogeneity of ovarian cancer.Arch Gynecol Obstet 2014;289:237–9.

Figure 5.Toxin-boundRL21Akills HLA-A2þovarian cancer cells. Cellswere treatedwithdilutions of either bare RL21A in the presence of a constant concentration ofcytotoxic secondary (A) or RL21A-deB (B) for 72 hours. Live cells weredetected by MTS assay. Data are mean� SD from at least three independentexperiments with at least 3 replicate wells per experiment. Data are fit withsigmoidal 4-parameter logistics curves (A2780-A2 and FHIOSE) or best-fitlines (A2780).






10. Han LY, Fletcher MS, Urbauer DL, Mueller P, Landen CN, Kamat AA, et al.HLA class I antigen processing machinery component expression andintratumoral T-Cell infiltrate as independent prognosticmarkers in ovariancarcinoma. Clin Cancer Res 2008;14:3372–9.

11. Zhang L, Conejo-Garcia JR, KatsarosD, Gimotty PA,MassobrioM, RegnaniG, et al. Intratumoral T cells, recurrence, and survival in epithelial ovariancancer. N Engl J Med 2003;348:203–13.

12. Hwang WT, Adams SF, Tahirovic E, Hagemann IS, Coukos G. Prognosticsignificance of tumor-infiltrating T cells in ovarian cancer: a meta-analysis.Gynecol Oncol 2012;124:192–8.

13. Verma B, Neethling FA, Caseltine S, Fabrizio G, Largo S, Duty JA, et al. TCRmimicmonoclonal antibody targets a specific peptide/HLA class I complexand significantly impedes tumor growth in vivo using breast cancermodels. J Immunol 2010;184:2156–65.

14. Jain R, Rawat A, Verma B, Markiewski MM, Weidanz JA. Antitumoractivity of a monoclonal antibody targeting major histocompatibilitycomplex class I-Her2 peptide complexes. J Natl Cancer Inst 2013;105:202–18.

15. Hagemann T, Wilson J, Kulbe H, Li NF, Leinster DA, Charles K, et al.Macrophages induce invasiveness of epithelial cancer cells via NF-kappa Band JNK. J Immunol 2005;175:1197–205.

16. Agarwal R, Whang DH, Alvero AB, Visintin I, Lai Y, Segal EA, et al.Macrophage migration inhibitory factor expression in ovarian cancer. AmJ Obstet Gynecol 2007;196:348 e1–5.

17. Hagemann T, Robinson SC, Thompson RG, Charles K, Kulbe H, Balkwill FR.Ovarian cancer cell-derived migration inhibitory factor enhances tumorgrowth, progression, and angiogenesis.MolCancer Ther 2007;6:1993–2002.

18. Krockenberger M, Dombrowski Y, Weidler C, Ossadnik M, Honig A,Hausler S, et al. Macrophage migration inhibitory factor contributes tothe immune escape of ovarian cancer by down-regulating NKG2D.J Immunol 2008;180:7338–48.

19. Kulbe H, Thompson R, Wilson JL, Robinson S, Hagemann T, Fatah R, et al.The inflammatory cytokine tumor necrosis factor-alpha generates anautocrine tumor-promoting network in epithelial ovarian cancer cells.Cancer Res 2007;67:585–92.

20. Alvero AB, Chen R, Fu HH, Montagna M, Schwartz PE, Rutherford T, et al.Molecular phenotyping of human ovarian cancer stem cells unravels themechanisms for repair and chemoresistance. Cell Cycle 2009;8:158–66.

21. Mor G, Yin G, Chefetz I, Yang Y, Alvero A. Ovarian cancer stem cells andinflammation. Cancer Biol Ther 2011;11:708–13.

22. Xu L, Li Y, Sun H, Zhen X, Qiao C, Tian S, et al. Current developments ofmacrophage migration inhibitory factor (MIF) inhibitors. Drug DiscovToday 2013;18:592–600.

23. Hussain F, FreissmuthM, Volkel D, Thiele M, Douillard P, Antoine G, et al.Human anti-macrophage migration inhibitory factor antibodies inhibitgrowth of humanprostate cancer cells in vitro and in vivo.Mol Cancer Ther2013;12:1223–34.

24. Huang XH, Jian WH, Wu ZF, Zhao J, Wang H, Li W, et al. Small interferingRNA (siRNA)-mediated knockdown of macrophage migration inhibitoryfactor (MIF) suppressed cyclin D1 expression and hepatocellular carcino-ma cell proliferation. Oncotarget 2014;5:5570–80.

25. Hawkins OE, Vangundy RS, Eckerd AM, Bardet W, Buchli R, Weidanz JA,et al. Identification of breast cancer peptide epitopes presented by HLA-A�0201. J Proteome Res 2008;7:1445–57.

26. Hawkins O, Verma B, Lightfoot S, Jain R, Rawat A, McNair S, et al. An HLA-presented fragment of macrophage migration inhibitory factor is a ther-apeutic target for invasive breast cancer. J Immunol 2011;186:6607–16.

27. Kruk PA,GodwinAK,Hamilton TC, AuerspergN. Telomeric instability andreduced proliferative potential in ovarian surface epithelial cells fromwomen with a family history of ovarian cancer. Gynecol Oncol 1999;73:229–36.

28. Wei ML, Cresswell P. HLA-A2 molecules in an antigen-processingmutant cell contain signal sequence-derived peptides. Nature 1992;356:443–6.

29. Hickman HD, Luis AD, Buchli R, Few SR, Sathiamurthy M, VanGundy RS,et al. Toward a definition of self: proteomic evaluation of the class I peptiderepertoire. J Immunol 2004;172:2944–52.

30. Kim S, Li L, McMurtrey CP, Hildebrand WH, Weidanz JA, Gillanders WE,et al. Single-chain HLA-A2 MHC trimers that incorporate an immundo-minant peptide elicit protective T cell immunity against lethal West Nilevirus infection. J Immunol 2010;184:4423–30.

31. Buchli R, VanGundy RS, Hickman-Miller HD, Giberson CF, Bardet W,Hildebrand WH. Real-time measurement of in vitro peptide binding tosoluble HLA-A�0201 by fluorescence polarization. Biochemistry 2004;43:14852–63.

32. Bolognesi A, Polito L, Olivieri F, Valbonesi P, Barbieri L, Battelli MG, et al.New ribosome-inactivating proteins with polynucleotide:adenosine gly-cosidase and antiviral activities from Basella rubra L. and bougainvilleaspectabilis Willd. Planta 1997;203:422–9.

33. Cizeau J,GrenkowDM,Brown JG, Entwistle J,MacDonaldGC. Engineeringand biological characterization of VB6-845, an anti-EpCAM immunotoxincontaining a T-cell epitope-depleted variant of the plant toxin bouganin.J Immunother 2009;32:574–84.

34. Bolognesi A, Polito L, Tazzari PL, Lemoli RM, Lubelli C, Fogli M, et al. Invitro anti-tumour activity of anti-CD80 and anti-CD86 immunotoxinscontaining type 1 ribosome-inactivating proteins. Br J Haematol 2000;110:351–61.

35. Weinzierl AO, Lemmel C, Schoor O, Muller M, Kruger T, Wernet D, et al.Distorted relation between mRNA copy number and corresponding majorhistocompatibility complex ligand density on the cell surface. Mol CellProteomics 2007;6:102–13.

36. Milner E, Barnea E, Beer I, Admon A. The turnover kinetics of majorhistocompatibility complex peptides of human cancer cells. Mol CellProteomics 2006;5:357–65.

37. Kessler M, Fotopoulou C,Meyer T. Themolecular fingerprint of high gradeserous ovarian cancer reflects its fallopian tube origin. Int J Mol Sci2013;14:6571–96.

38. O'Shannessy DJ, Jackson SM, Twine NC, Hoffman BE, Dezso Z, Agoul-nik SI, et al. Gene expression analyses support fallopian tube epitheliumas the cell of origin of epithelial ovarian cancer. Int J Mol Sci2013;14:13687–703.

39. Nik NN, Vang R, Shih Ie M, Kurman RJ. Origin and pathogenesis of pelvic(ovarian, tubal, and primary peritoneal) serous carcinoma. Annu RevPathol 2014;9:27–45.

40. NorellH,CarlstenM,OhlumT,MalmbergKJ,MasucciG, SchedvinsK, et al.Frequent loss of HLA-A2 expression in metastasizing ovarian carcinomasassociated with genomic haplotype loss and HLA-A2-restricted HER-2/neu-specific immunity. Cancer Res 2006;66:6387–94.

41. Rolland P, Deen S, Scott I, Durrant L, Spendlove I. Human leukocyteantigen class I antigen expression is an independent prognostic factor inovarian cancer. Clin Cancer Res 2007;13:3591–6.

42. Vermeij R, Daemen T, de Bock GH, de Graeff P, Leffers N, Lambeck A, et al.Potential target antigens for a universal vaccine in epithelial ovarian cancer.Clin Dev Immunol 2010;2010.

43. Andersson E, Villabona L, Bergfeldt K, Carlson JW, Ferrone S, Kiessling R,et al. Correlation of HLA-A02� genotype and HLA class I antigen down-regulation with the prognosis of epithelial ovarian cancer. Cancer Immu-nol Immunother 2012;61:1243–53.

44. Stam NJ, Vroom TM, Peters PJ, Pastoors EB, Ploegh HL. HLA-A-and HLA-B-specific monoclonal antibodies reactive with free heavychains in western blots, in formalin-fixed, paraffin-embedded tissuesections and in cryo-immuno-electron microscopy. Int Immunol1990;2:113–25.

45. Kobayashi M, Chiba A, Izawa H, Yanagida E, Okamoto M, Shimodaira S,et al. The feasibility and clinical effects of dendritic cell-based immuno-therapy targeting synthesized peptides for recurrent ovarian cancer.J Ovarian Res 2014;7:48.

46. Mitchell RA, YaddanapudiK. Stromal-dependent tumor promotion byMIFfamily members. Cell Signal 2014;26:2969–78.

47. Panowksi S, Bhakta S, RaabH, Polakis P, Junutula JR. Site-specific antibodydrug conjugates for cancer therapy. mAbs 2014;6:34–45.

48. Awada A, Bozovic-Spasojevic I, Chow L. New therapies in HER2-positivebreast cancer: a major step towards a cure of the disease? Cancer Treat Rev2012;38:494–504.


Patterson et al.




2016;15:313-322. Published OnlineFirst December 30, 2015.Mol Cancer Ther Andrea M. Patterson, Saghar Kaabinejadian, Curtis P. McMurtrey, et al. Target for Ovarian CancerInhibitory Factor Is a Surface Biomarker and Potential Therapeutic

Presented Macrophage Migration−Human Leukocyte Antigen

Updated version

10.1158/1535-7163.MCT-15-0658doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2015/12/30/1535-7163.MCT-15-0658.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/15/2/313.full#ref-list-1

This article cites 46 articles, 14 of which you can access for free at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://mct.aacrjournals.org/content/15/2/313To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-15-0658http://mct.aacrjournals.org/content/suppl/2015/12/30/1535-7163.MCT-15-0658.DC1http://mct.aacrjournals.org/content/15/2/313.full#ref-list-1http://mct.aacrjournals.org/cgi/alertsmailto:[email protected]://mct.aacrjournals.org/content/15/2/313http://mct.aacrjournals.org/

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages false /GrayImageMinResolution 200 /GrayImageMinResolutionPolicy /Warning /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages false /MonoImageMinResolution 600 /MonoImageMinResolutionPolicy /Warning /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 900 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MarksOffset 18 /MarksWeight 0.250000 /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /NA /PageMarksFile /RomanDefault /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /LeaveUntagged /UseDocumentBleed false >> > ]>> setdistillerparams> setpagedevice

Documents

Human Leukocyte Antigen Presented Macrophage Migration ......Companion Diagnostics and Cancer Biomarkers Human Leukocyte Antigen–Presented Macrophage Migration Inhibitory Factor