The Prostate 62:14 ^26 (2005)

AntinuclearAutoantibodies in ProstateCancer:Immunity to LEDGF/p75, a Survival ProteinHighly Expressedin ProstateTumors and

CleavedDuringApoptosis

Tracy Daniels,1 Jianying Zhang,2 Israel Gutierrez,3 Max L. Elliot,4 Brian Yamada,5

Mary Jo Heeb,6 Shaun M. Sheets,1 Xiwei Wu,1 and Carlos A. Casiano1,3*1Departmentof BiochemistryandMicrobiologyandCenter forMolecular BiologyandGeneTherapy,

Loma LindaUniversity SchoolofMedicine, Loma Linda,California2Departmentof Biology,Universityof Texas,El Paso,Texas

3DepartmentofMedicine, Loma LindaUniversity SchoolofMedicine, Loma Linda,California4DepartmentofUrologyand Pathology,The Scripps Clinic, La Jolla,California

5Departmentof Surgery,DivisionofUrology, Loma LindaUniversity SchoolofMedicine, Loma Linda,California6DepartmentofMolecularandExperimentalMedicine,The Scripps Research Institute, La Jolla,California

BACKGROUND. Cancer patients produce autoantibodies to self-proteins called tumor-associated antigens (TAA). These autoantibodies represent potentially valuable tools foridentifying novel biomarkers and therapeutic targets. This studywas designed to identify TAAin prostate cancer (PCa).METHODS. Serum autoantibodies to the survival protein lens epithelium-derived growthfactor p75 (LEDGF/p75) were detected by immunofluorescence microscopy, ELISA, andimmunoblotting. Expression of LEDGF/p75 in prostate cells and tumors was evaluated byimmunoblotting or immunohistochemistry. Apoptotic cleavage of LEDGF/p75 was detectedby immunoblotting.RESULTS. Anti-LEDGF/p75 autoantibodies were detected by ELISA in 18.4% of PCa patientsand 5.5% of matched controls (P< 0.001) but not in patients with benign prostatic hyperplasia(BPH). LEDGF/p75 expression was detected in 93% of prostate tumors but not in normalprostate. Strong expression of the protein was observed in 61% of prostate tumors. Moderate tohigh expression was also detected in BPH tissue. Cleavage of LEDGF/p75 was detected inapoptotic prostate cells.CONCLUSIONS. The high expression of LEDGF/p75 in prostate tumors and BPH could beinduced by inflammation and oxidative stress. LEDGF/p75 cleavage fragments generatedduring prostate tumor cell death might trigger autoantibodies under inflammatory conditionsin certain patients. Prostate 62: 14–26, 2005. # 2004 Wiley-Liss, Inc.

KEY WORDS: apoptosis; autoantibodies; inflammation; oxidative stress; LEDGF/p75;prostate cancer; tumor-associated antigens

Grant sponsor: National Institutes of Health; Grant numbers:AI44088, CA59979, RR00833; Grant sponsor: National MedicalTechnology TestBed, Inc.; Grant number: DAMD17-97-2-7016; Grantsponsor: Loma Linda University School of Medicine; Grant sponsor:General Clinical Research Center of The Scripps Research Institute;Grant sponsor: Hedco Foundation (support for Microscopy andImaging Facilities).

*Correspondence to: Carlos A. Casiano, Loma Linda UniversitySchool ofMedicine, Center forMolecular Biology andGene Therapy,132 Mortensen Hall, 11085 Campus Street, Loma Linda, CA 92350.E-mail: ccasiano@som.llu.eduReceived 16 December 2003; Accepted 26 March 2004DOI 10.1002/pros.20112Published online 26 April 2004 in Wiley InterScience(www.interscience.wiley.com).

� 2004 Wiley-Liss, Inc.

INTRODUCTION

The serum autoantibody repertoire from cancerpatients is currently being actively exploited for theidentification of tumor-associated antigens (TAA) andthe design of antigen panels ormicroarrays [1–3]. Sucharrays would facilitate autoantibody profiling andpotentially aid in the serological diagnosis and prog-nosis of cancer. The factors leading to autoantibodyproduction in cancer patients are not clear, althoughemerging data indicate that TAA are autologousintracellular proteins (e.g., survivin and p53) thatpresent altered regulation, structure, or function intumors [4–7]. Hence, cancer-associated autoantibodieshave been considered as reporters identifying aberrantcellular mechanisms underlying malignancy [5].

Autoantibody responses to intracellular proteinshave been extensively investigated in a wide variety ofcancers [1–7]. However, information about theseresponses in prostate cancer (PCa) is still limited. Earlystudies demonstrated the presence of antinuclearautoantibodies (ANA) in PCa patients [8,9] but theautoantibodieswere also found in patientswith benignprostatic hyperplasia (BPH). This, along with the factthat ANA had not yet been widely associated withcancer at the time, led to the conclusion that in PCathese autoantibodies might be representative of non-specific markers pointing to other immunologicalprocesses [8,9].

In recent years, several groups have exploited theserum autoantibody repertoire from PCa patients toidentify candidate TAA. The growing list of theseantigens include 5a-reductase [10], p53 [11], prosta-somes [12], glucose-regulated protein-78 kDa (GRP78)[13], MUC1 [14], PARIS-1 [15], and several cancer/testis antigens [16]. Zhang et al. [2] demonstratedrecently the presence of serum autoantibodies to apanel of known TAA in various human cancers,including PCa. In that study, autoantibody reactivityto individual TAA rarely exceeded 20% in the cancerpatient populations (as compared to less than 5% innormal human sera (NHS). However, successiveaddition of TAA to a mini-array of seven antigensresulted in a dramatic increase in the number ofpositive antibody reactions in cancer patients but notin healthy controls. Interestingly, cancers of theprostate, breast, and lung showed distinct profiles ofautoantibody reactivity in which specific TAA werepreferentially targeted. These results strongly sug-gested that uniquely constituted antigen panels ormini-arrays might provide an alternative approach fordiscriminating autoantibody reactivity between cancerpatients and control individuals and distinguishingbetween some types of cancer. The identification ofadditional antigens targeted by the immune system in

PCa patients is, therefore, essential for designingspecific antigen arrays for autoantibody profiling inPCa cancer populations and for defining targets ofnovel strategies for the treatment of this cancer type.

To advance our identification of potential TAAtargeted by circulating autoantibodies in PCa patients,we undertook a comprehensive analysis of the serumANA repertoire in this cancer type. We observed thatwhile the general frequency and titers of ANA in PCapatients were relatively similar to those in matchedcontrols, significant differences could be detectedbetween the two groups in the autoantibody responseto an individual autoantigen, the lens epithelium-derived growth factor p75 (LEDGF/p75). In addition,we detected high expression of this antigen in prostatetumors, particularly advanced stage tumors, but not innormal prostate. Cleavage of LEDGF/p75 was alsodetected in apoptotic prostate cells. LEDGF/p75, alsoknown as transcription co-activator p75 [17,18] anddense fine speckled protein of 70 kDa (DFS70) [19], is asurvival protein that protects mammalian cells fromoxidative stress-induced cell death by upregulating theexpression of anti-apoptotic heat shock proteins (Hsps)[18,20]. We demonstrated recently that this pro-survival function is abrogated during apoptosis bycaspase-induced cleavage of the protein [21]. Theautoantibody response to LEDGF/p75 in certain PCapatients suggests that the immunogenicity of thisproteinmight be increased in prostate tumors, possiblyas a consequence of increased expression combinedwith proteolytic cleavage in tumor cells dying under aproinflammatory environment.

MATERIALSANDMETHODS

SerumSamples

Serum samples were obtained with informed con-sent from randomly selected patients with PCa(n¼ 207) and a control group of age-matched malepatients with no clinical evidence of PCa (n¼ 166) seenat Loma Linda University Medical Center and FacultyMedical Offices. The average age was 71.9 years old(range 46–100) for the PCa group and 64.9 years old(range 46–100) for the matched control group. It hasbeen established in autopsy studies that silent PCa isoften present in elderlymales [22]. Since this studywasoriginally designed to evaluate the serum antinuclearautoantibody repertoire in PCa, three additional con-trol groups less likely to have PCawere also selected forthe study. The first group consisted of healthy blooddonors (n¼ 82) with an average age of 31.5 (range 18–59). Sera from these donors, designated NHS, wereobtained from the serum bank of the W.M. KeckAutoimmune Disease Center of The Scripps ResearchInstitute, La Jolla, and were used to establish ELISA

Immunity to LEDGF/p75 in Prostate Cancer 15

cut-off values in previous studies of cancer-associatedautoantibodies [2,23]. The second group consisted of arandom selection of 44male blood donors from the SanBernardino Blood Bank, with an average age of 47.0(range 20–76), andwas used inANAscreening studies.The third group consisted of sera drawn from 40 pa-tients with BPH from the General Clinical ResearchCenter of The Scripps Research Institute, La Jolla, withan average age of 70.3 (range 51–89). This study wasapproved by the Institutional Human Subject ReviewBoards of Loma Linda University and The ScrippsResearch Institute.

Cell Lines

Cell lines were obtained from the American TypeCulture Collection (Rockville, MD), with the exceptionof PrEC (Cambrex, Walkersville, MD) and 55T and 41T(BRFF, Jamsville, MD). Cells were grown in 5% CO2 at378C using the media recommended by the suppliers.

Preparation of RecombinantHis-TaggedLEDGF/p75

ApET-28a plasmid (Novagen,Madison,WI), encod-ing full-length LEDGF/p75 with an NH2-terminal6xHis tag was transformed into BL21-Codon Pluscompetent Escherichia coli (Stratagene, La Jolla, CA).Bacterial cultures induced to overexpress LEDGF/p75with isopropyl-beta D-thiogalactopyranoside (IPTG)were centrifuged and pellets were resuspended inbinding buffer containing 20 mM sodium phosphatepH 7.4, 0.5 M NaCl, 0.02% NaN3, and the CompleteTM

protease inhibitor cocktail (Roche, Indianapolis, IN).The suspension was sonicated, centrifuged, and thesupernatant was passed through a HiTrap Chelatingcolumn using the AKTAFPLC (Amersham Biosciences,Piscataway, NJ). Bound LEDGF/p75 was eluted withbinding buffer containing 0.5 M imidazole. The purityand integrity of recombinant LEDGF/p75was assessedby SDS–PAGE and Coomassie Blue staining (data notshown). Only preparations containing intact LEDGF/p75 were used for ELISA analysis after determiningprotein concentrations using the Eppendorf Biophot-ometer (Brinkman, Westborg, NY).

Indirect Immunofluorescence (IIF)Microscopy

Prostate cells grownon coverslipswerewashedwithPBS, fixed, and permeabilized in methanol/acetone(3:1 v/v), and stained with a highly specific humananti-LEDGF/p75 serum from a PCa patient diluted at1:200. Fluorescein-5-isothiocyanate (FITC)-conjugatedgoat anti-human IgG antibody (Caltag Laboratories,San Francisco, CA) was used as secondary detectingreagent at a 1:100 dilution. Cells were counterstained

with 40,6-diamidino-2-phenylindole (DAPI) for DNAvisualization and examined under a fluorescence micro-scope. Pre-fixed HEp-2 cell slides (Bion, Park Ridge, IL)were used to detect serum ANA following the stainingprocedure indicated above. Sera were initially tested at a1:100dilutionandconsideredpositive forANAonly if thenuclear or cytoplasmic fluorescence in the HEp-2 cellswas significantly higher than the fluorescence of cellsincubatedwithNHSused at a 1:100 dilution. Discrimina-tionbetweenpositive andnegative serawas facilitatedbya Spot Imaging System (Diagnostic Instruments, SterlingHeights,MI),which allowed the simultaneousdisplay ona computer screen of images from NHS, positive controlsera, and experimental sera. All images were acquiredunder identical conditions. Ambiguous results wereconsidered negative.

Immunoblotting

Whole cell lysates were prepared as indicatedpreviously [24] and the protein concentration in thelysates was determined using the BioRad DC ProteinAssay Kit (BioRad Laboratories, Hercules, CA) toensure equal loading of proteins per lane in sodiumdodecyl sulfate–polyacrylamide gels. Electrophoresisand immunoblottingwere performed asdescribed [24].Patient sera were used at a dilution of 1:100.

ELISA

ELISA procedures were performed essentially asindicated by Zhang et al. [2]. Briefly, purified recombi-nant LEDGF/p75 was diluted in phosphate bufferedsaline (PBS) to a final concentration of 0.5 mg/ml and200 ml of this dilution were used to coat Immulon-2Flatbottom microtiter plates (Fisher Scientific, Tustin,CA) for 16 hr at 48C. After removal of the proteinsolution, plates were blocked with 0.1% gelatin in PBSfor 2 hr at 48C. The plates were washed four times withPBS/0.05% Tween-20 (PBST). Patient sera diluted at1:200 in serum diluent solution (1 mg/ml gelatin,0.75 mg/ml bovine gamma globulin, and 1 mg/mlbovine serum albumin in PBST) were then added toeachwell. After 1.5 hr at room temperature, plateswerewashed four times with PBST. Each well was thenincubated for 1.5 hr with horseradish peroxidase(HRP)-conjugated goat anti-human IgG (Caltag Labs,San Francisco, CA) at a 1:4,000 dilution. The plateswerewashed four times with PBS and developed byincubating each well with 200 ml peroxidase substratesolution (1mg/ml 2,2-azino-bis[3-ethylbenzthiazoline-6-sulfonic acid] with 0.005% hydrogen peroxide incitrate buffer, pH 4.6) for 2 hr. Each serum sample wastested in triplicate and the average absorbance at490 nm was used for data analysis. The cutoff valuedesignating positive reactions was calculated as three

16 Daniels et al.

standard deviations (SD) above the mean absorbancevalue for NHS (n¼ 82 for ELISA of PCa sera and n¼ 40for ELISA of BPH sera). Patient sera and NHS weretested simultaneously in triplicate in at least twoindependent experiments. Each ELISA experimentincluded a rabbit antibody raised against the carboxylterminus of LEDGF/p75 as positive control serum.For pre-absorbtion experiments, diluted patient sera(1:200) were incubated with 15 mg/ml of recombinantLEDGF/p75 for 1 hr at 378C prior to testing by ELISA.

Immunohistochemistry

Deparaffinized and rehydrated human PCa HistoTM-Array slides (Imgenex, San Diego, CA) and BPH tissue(ScrippsClinic, La Jolla, CA)mounted on Superfrost Plusslides (Fisher Scientific, Pittsburgh, PA) were stainedusing a Biogenic i6000 auto stainer (Biogenex Corpora-tion, San Ramon, CA) following the manufacturer’sstandardized immunohistochemistry protocol. Briefly,the sections were immersed in the Citra-Plus antigenretrieval solution (BiogenexCorporation). Antigen retrie-valwasperformedbymicrowaving the slides for 2min at100% power followed by 10 min at 20% power. Slideswere then cooled in the antigen retrieval solution for20 min. Non-specific peroxidase activity was blocked for15minwith 3%hydrogen peroxide in 10%methanol andnon-specific protein binding sites were blocked withPower Block� universal blocking reagent (BiogenexCorporation) for 10 min. Sections were incubated for1 hr with a primary rabbit anti-LEDGF/p75 antibody(1:500), rinsed in PBS three times, then incubatedwith the Multi-link� (Biogenex Corporation) biotiny-lated secondary antibody for 20 min followed byincubation with a streptavidin-coupled peroxidaseSupersensitive Label� (Biogenex Corporation) for20min.Theantibody-localizedantigenwas thendetectedby a 5-min peroxidase activation of a water-insoluble3-amino-9-ethylcarbazole (AEC) chromagen (RomulinAEC-Biocare Medical, Walnut Creek, CA) that forms areddish-orange precipitate. Sections were then counter-stained lightly with hematoxylin (Sigma) and mountedwith Permount (Fisher Scientific, Fair Lawn, NJ).

Induction of Apoptosis

Apoptosiswas induced in prostate cells by exposureto 4 mM staurosporine (STS) for up to 24 hr or byexposure to 100 ng/ml actinomycin D (Act D) for 4 hrfollowed by 100 ng/ml TRAIL (R&D Systems, Min-neapolis, MN) for up to 24 hr.

Statistical Analysis

w2 analysis was used to determine the significance ofdifferences in ANA frequencies in the PCa group andthe control groups, and to correlate LEDGF/p75staining in tumor tissue with tumor stage. For theELISA data, an unpaired t-test (two tailed) was used,while for the pre-absorbtion experiments, a pairedanalysis (two tailed) was performed.

RESULTS

Analysis of the CirculatingANARepertoirein PCaby IIFMicroscopy

Using IIF microscopy in HEp-2 slides, we screenedsera from 207 patients with clinically diagnosed PCa,166 age-matched men with no clinical evidence of PCa(based on medical history), and 44 randomly selectedmale blood bank donors for the presence of circulatingANA. The ANA frequency was 29% for PCa patients,24% for the matched controls, and 4.5% for the bloodbank donors (Table I). ANA frequency (based onpositive reaction at 1:100 dilution) was significantlyincreased in both the PCa group and thematched controlgroup when compared to the blood bank donors(P¼ 0.002 and P¼ 0.02, respectively). The average ANAtiters for the three groups were 1:450, 1:344, and 1:160,respectively. While the general frequency and titers ofANA in the PCa and matched control groups were notsignificantly different, it was noted that differences inspecific antibody staining patterns could be detectedbetween the twogroups.A stainingpattern characterizedby dense fine speckles in interphasic nuclei and brightstaining ofmitotic chromosomeswas predominant in thePCa group (Table I, Fig. 1). This staining pattern,

TABLE I. ANAPatterns and Titers in Sera FromPCaPatients

Numbertested

Average age(range)

ANAtiter

ANAþ(%)

ANA pattern (average titer)

DFS ON C CS M NL

PCa 207 71.9 (46–100) 450 60 (29) 29 (367) 14 (229) 10 (312) 2 (240) 3 (2,240) 2 (1,360)Matched control group 166 64.9 (46–100) 344 40 (24) 7 (594) 23 (285) 6 (227) 1 (320) 2 (480) 1 (320)Male blood bank donors 44 47.0 (20–76) 160 2 (4.5) 1 (160) 0 (0) 0 (0) 1 (160) 0 (0) 0 (0)

ANA, antinuclear autoantibodies; Pca, prostate cancer; DFS, dense fine speckles; ON, other non-DFS diffuse or homogeneous nuclear; C,cytoplasmic including mitochondrial; CS, Course Speckled; M, mitotic apparatus including centromere and NuMA; NL, nucleolar.

Immunity to LEDGF/p75 in Prostate Cancer 17

characteristic of autoantibodies to DFS70 (an alterna-tive name for LEDGF/p75) [19], was present in 14%(29/207) of the PCa sera, compared to 4.2% (7/166) insera from the matched control group, and 2.3% in serafrom blood bank donors (Table I). Other ANA stainingpatterns were also observed in the three groups(Table I).

Detection of LEDGF/p75Autoantibodiesin PCa Seraby ELISA

To explore the possibility that LEDGF/p75 could bethe target of an autoantibody response in some patientswith PCa, sera from PCa patients and the matchedcontrol individuals were tested by ELISA against puri-fied recombinant LEDGF/p75. This analysis showedthat 18.4% (38/206) of the PCa patients producedautoantibodies to LEDGF/p75, compared to 5.5% (9/164) of the matched controls (Fig. 2A). The meanabsorbance value for the whole PCa group was0.213� 0.02 whereas the mean value for the matchedcontrol group was 0.047� 0.008 (P< 0.001). The cutofffor positive antibody reactivity against LEDGF/p75

Fig. 1. Detection of autoantibodies to lens epithelium-derivedgrowth factor p75 (LEDGF/p75) by indirect immunofluorescence(IIF) microscopy in sera from patients with prostate cancer (PCa).The stainingpatterninHEp-2cells of areferencehumanserumcon-taining autoantibodies to LEDGF/p75 [19] and a rabbit anti-LEDGF/p75antibodyare shown.Thispatternconsists ofdense fine specklesininterphasenucleiandbrightstainingofmitoticchromosomes.Tworepresentative sera frompatientswith PCa (P096 and P119) displaythe same LEDGF/p75 pattern. All cells were counterstained withDAPI to confirmthenuclear localizationofLEDGF/p75.

Fig. 2. Detection of anti-LEDGF/p75 reactivityby ELISA.A: Anti-LEDGF/p75 autoantibodies were detected by ELISA in sera frompatients with PCa (n¼ 206) and a gender- and age-matched controlgroup(n¼164).ThehorizontallineindicatesthecutoffvalueforpositiveantibodyreactivityagainstLEDGF/p75(>0.268),whichwasdefinedasanabsorbancegreater than three standarddeviations above themeanvaluefor82normalhumansera(NHS).B:Anti-LEDGF/p75autoantibo-diesinserafrompatientswithBPH(n¼ 40)andNHS(n¼ 40).Thecut-off value for positive antibody reactivity against LEDGF/p75 (>0.103)was defined as an absorbance greater than three standard deviationsabove themeanvalue for the 40NHS.C: Analysis of the specificityofanti-LEDGF/p75 reactivity in ELISA by blocking selected patient serawith soluble recombinant protein. Five PCa sera (P096, P020, P010,P021,P082) and twomatchedcontrol sera (C003andC005)werepre-incubatedwith (blackbars) andwithout (whitebars)purifiedrecombi-nantLEDGF/p75andthentestedbyELISA.Significanceofthereduction(*)wasP< 0.001usingapaired, two-tailedStudent’st-test.Thereactiv-ity of sera from PCa patients and thematched controls was assessedsimultaneously.Alltheresultsshownaremeanvaluesof threedetermi-nationsinatleasttwoindependentexperiments.

18 Daniels et al.

was 0.268, which was defined as an absorbance greaterthan three standard deviations above the mean valuefor 82 NHS obtained from healthy donors. Thisrelatively stringent cutoff criterion has been used inprevious studies on autoantibody responses in cancer[2,4,23]. The mean absorbance value of NHS was0.057� 0.070. Only two NHS had values above thecut-off (0.399 and 0.298, respectively). We used theNHS, rather than the gender- and age-matched controlgroup, to determine the cutoff for positivity to facilitatea comparison of the frequency and levels of anti-LEDGF/p75 antibodies in PCa patients with those ofage-matched men without clinically apparent PCa. Nosignificant differences were observed in the frequencyof autoantibodies to LEDGF/p75 inNHS and BPH sera(Fig. 2B).

To determine the specificity of the anti-LEDGF/p75antibody reactivity, representative patient sera (fivePCa and two matched control) were pre-incubatedwith recombinant LEDGF/p75 and then tested byELISA. Non-pre-absorbed sera were tested simulta-neously with matching pre-absorbed sera. The specifi-city of the anti-LEDGF/p75 reactivity was evidencedby the significant decrease of absorbance values afterpre-absorption (Fig. 2C).

Detection of LEDGF/p75Autoantibodiesin PCa Seraby Immunoblotting

To confirm that LEDGF/p75 was a target of anautoantibody response in some patients with PCa, weperformed immunoblotting analysis of the ELISA andIIF positive sera. Figure 3 shows representative serareacting in immunoblots of whole LNCaP cell lysateswith a protein of approximately 75 kDa that hadidentical migration as the LEDGF/p75 protein recog-nized by a reference antibody. The immunofluores-cence staining patterns shown in Figure 1 and thereactivity with a band around the 75 kDa region aretypical of serum autoantibodies to LEDGF/p75 [19].

In the immunoblotting analysis, eight of the 38 PCasera (27.7%) positive by ELISA reacted strongly with a

75 kDa band in the lysates (examples provided in Fig. 3:sera P001, P002, and P096). Intriguingly, none of thenine matched control sera that were positive by ELISAreacted with this band (data not shown). We alsoidentified eight additional PCa sera with IIF stainingpatterns typical of that of anti-LEDGF/p75 antibodies,with ELISA reactivity just below the cut-off, whichwere positive by immunoblotting (see Fig. 3: sera P019,P051, P072, P088, P093, P115, P119, and P180). Two serafrom the matched control group displaying the densefine speckled IIF pattern and poorly reactive by ELISA,also reacted with a 75 kDa band (see Fig. 2: sera C069and C134). In summary, 46 of 206 PCa sera (22.3%)reacted with LEDGF/p75 by either ELISA or byimmunoblotting against LNCaP lysates, in contrast to11 of 164 (6.7%)matched controls. Upon examining theclinical history of themajority of the patients producingthe anti-LEDGF/p75 antibody,we foundno correlationbetween the presence of this antibody and PSA levels,Gleason score, tumor stage, or therapeutic regimen(data not shown).

Expression of LEDGF/p75 Proteinin Prostate Cell Lines

Previous studies on LEDGF/p75 have shown itsexpression in various non-tumormammalian cell typessuch as lens epithelial cells, fibroblasts, and keratino-cytes [20] and in a few tumor cell lines (e.g., Jurkat,HeLa, and HepG2) [21]. However, to date there are nopublished studies addressing whether this protein iswidely expressed in tumor cells. Since the data pre-sented above indicated that LEDGF/p75 is the target ofhumoral immunity in a subset of patients with PCa, wesought todeterminewhether theprotein is expressed inprostate tumor cells and in other tumor cell types.Using a high titer, relatively monospecific anti-LEDGF/p75 serum from patient P096, we detectedexpression of LEDGF/p75 by immunoblotting in apanel of prostate cell lines that included six tumor(DU145, PC3, LNCaP, 41T, 22Rv1, RWPE-2), twotransformed non-tumor (RWPE-1 and Pwr-1E), onetransformed BPH (55T), and one primary normalprostate (PrEC) (Fig. 4A). The lowest LEDGF/p75expression appeared to be in the normal PrEC cell line,suggesting the possibility that the protein mightbe overexpressed in tumor or transformed cells.Expression of LEDGF/p75 was also detected in othernon-prostate tumor cell lines (Fig. 4B), with lowerexpression observed in liver HepG2 tumor cells.

Using IIF microscopy, we observed strong nuclearstaining associated with moderate cytoplasmicstaining in LNCaP (androgen-dependent) and PC3(androgen-independent) prostate tumor cells (Fig. 5).Relatively weaker staining was observed in PrEC

Fig. 3. Immunoblotofrepresentative sera frompatientswithPca(P) and gender- and age-matched control individuals (C) that dis-played the immunofluorescence staining pattern shown in Figure1.Note that these sera reacted with a protein of approximately75 kDa with migration identical to the LEDGF/p75 protein recog-nizedby a referencehuman serum [19].Theblot is representative ofatleast threeindependentexperiments.

Immunity to LEDGF/p75 in Prostate Cancer 19

normal prostate cells, consistent with the immunoblot-ting results (Fig. 4). The nuclear localization of LEDGF/p75 has been documented previously [19].

Expression of LEDGF/p75 Protein in ProstateTissue

To explore the possibility that LEDGF/p75might beoverexpressed in prostate tumors, an expressionanalysis of the protein in prostate tissuewas performedby immunohistochemistry. For these experiments, we

used human PCa HistoTM-Array slides (Imgenex).These slides contain 56 prostate tumor specimens andthree normal prostate specimenswithin the same tissuearray. Clinical information for each specimen, pro-vided by Imgenex, is summarized in Table II. It shouldbe noted that the Gleason score provided by Imgenexwas taken directly from the pathology report. Thisscore is the sum of the two most frequent histologicalpatterns of the entire prostate. Thus, the tissue sampleon the array may not necessarily be representative ofthat score. For this reason, a field Gleason score wasassigned to most tissue samples. However, a few of thetissues could not be scored since a hematoxylin andeosin stained array was not available.

The HistoTM-Array slides were stained with either arabbit anti-LEDGF/p75 serum or the rabbit pre-immune serumas an irrelevant antibody control. Stain-ing intensity was scored as strong, weak, or negative.The three normal tissues (#57, 58, 59) showed very littleor no staining (Table II and Fig. 6A), as determined bythe similar staining intensity of the anti-LEDGF/p75and pre-immune sera. One tumor tissue (#13) showedstrong positive staining with both the pre-immuneserum and the LEDGF/p75 antibody and thus was notincluded in the analysis. Tissue sample #35 was alsoexcludedbecause itwasofpoorquality.Of the54cancerspecimens, 50 (93%) expressed theprotein. Thirty-three(61%) tumor samples showed strong staining, 17 (31%)showed weak staining, and four (7%) were negative(Tables II and III; see representative stained specimensin Fig. 6A,B). Most of the tumors showing highexpression of LEDGF/p75 were stage III and IVtumors; however, because the number of tumors instages I–III provided in the tissue array was relativelysmall, it was not possible to establish a statisticallysignificant correlation between high LEDGF/p75 ex-pression and tumor stage. Of the three tumor speci-mens that were negative, one was from a patientdiagnosed with embryonal rhabdomyosarcoma, a softtissue tumor of striated muscle that can affect anyorgan. Interestingly, LEDGF/p75was also found in thetumor samples to be highly expressed in morphologi-cally normal tissue within the tumor microenviron-ment (Fig. 6C). Moderate to high expression of theprotein was also observed in all 12 tissue specimensfrom patients with BPH evaluated for the presence ofthis protein (Fig. 6D). The anti-LEDGF/p75 staining inprostate tumors and BPH tissue was mostly confinedto the nucleus, although cytoplasmic staining was alsoevident in many cells.

Cleavage of LEDGF/p75 inApoptotic Prostate Cells

Dying cells (both apoptotic and necrotic) areconsidered to be reservoirs of potentially immunosti-mulatory forms of intracellular antigens [24–28]. The

Fig. 4. ExpressionofLEDGF/p75inprostateandother tumorcelllinesdetectedbyimmunoblotting.A:ExpressionofLEDGF/p75 inapanel of prostate cell lines (tumor:DU145, PC3, LNCaP, 41T, 22Rv1,RWPE-2;transformednon-tumor:RWPE-1andPwr-1E;transformedbenignprostatichyperplasia(BPH):55T;andprimarynormal,PrEC).Twentymicrogramsof totalcellularproteinwereloadedperlane.b-Actinwasusedasa sampleloadingcontrol.B:ExpressionofLEDGF/p75 in the tumorcell lines Jurkat,A293,HL60,Colo320, andHepG2,and in FDC-P1myeloid hematopoietic progenitor cells. LEDGF/p75was detected inwhole cell lysates blottedonto nitrocelluloseusingserum from PCa patient P096. Blots are representative of at leasttwoindependentexperiments.

Fig. 5. ExpressionofLEDGF/p75inprostatecelllinesdetectedbyIIFmicroscopy.LEDGF/p75 is expressedpredominantly in thenucleiofLNCaP,PC3, andPrECcell lines.CorrespondingDAPI images areprovided. Immunostaining was performed using serum from PCapatientP096.Imagesarerepresentative of atleast twoindependentexperiments.

20 Daniels et al.

TABLE II. Clinical InformationAssociatedWith the ProstateTissues Providedonthe ImgenexHisto-TM-Array Slides

Sample Age Diagnosis GleasonField

Gleason TNM StageLEDGF/p75staining

1 74 Adenocarcinoma 7 3 T2aN0M0 II N2 88 Adenocarcinoma 10 5 T4NxMb IV S3 76 Adenocarcinoma 5 3 TxNxM1c IV S4 70 Adenocarcinoma 10 4 T3aN0M1b IV S5 64 Adenocarcinoma 7 3 T2N0M0 II W6 69 Adenocarcinoma 5 3 T2bN0M0 II S7 80 Adenocarcinoma 7 4 TxN0M0 II S8 74 Adenocarcinoma 6 3 T3aN0M0 III S9 56 Adenocarcinoma 4 3 T2N0M0 II S10 87 Adenocarcinoma 9 4 TxNxM1b IV S11 65 Adenocarcinoma 8 ? TxNxM1b IV N12 51 Adenocarcinoma 10 5 T3bN1M0 IV S13 81 Adenocarcinoma 9 5 S*14 70 Adenocarcinoma 8 5 M1 IV W15 71 Adenocarcinoma 6 ? T2bN0M0 II W16 64 Adenocarcinoma 9 4 M1 IV W17 72 Adenocarcinoma 9 4 T4N0M0 IV S18 67 Adenocarcinoma 3 3 T2N0M0 II W19 59 Adenocarcinoma 7 4 T2N1M0 IV N20 72 Adenocarcinoma 7 4 T1NxM0 I W21 78 Adenocarcinoma 7 3 T2N0M0 II W22 73 Adenocarcinoma 9 4 M1b IV W23 73 Adenocarcinoma 10 4 T4NxM1 IV S24 63 Adenocarcinoma 7 4 M1b IV S25 75 Adenocarcinoma 9 4 S26 57 Adenocarcinoma 10 4 T4NxM1c IV S27 60 Adenocarcinoma 10 3 TxN1M1 IV S28 78 Adenocarcinoma 4 3 T3N0M0 III S29 61 Adenocarcinoma 9 4 TxN1M0 IV S30 69 Adenocarcinoma 9 4 T2N0M0 II W31 63 Adenocarcinoma 9 5 T4NxM1 IV S32 60 Adenocarcinoma 8 4 T4NxM1c IV S33 61 Adenocarcinoma 9 4 TxN1M0 IV W34 71 Adenocarcinoma 8 3 S35 64 Adenocarcinoma 9 4 U36 61 Adenocarcinoma 9 5 TxN1M0 IV W37 69 Adenocarcinoma 9 5 T2N0M0 II W38 61 Adenocarcinoma 9 4 T4NxM1 IV S39 68 Adenocarcinoma 9 3 T3N1M0 IV W40 44 Adenocarcinoma 9 3 T2N1M0 IV W41 77 Adenocarcinoma 7 4 T2N0M0 II S42 74 Adenocarcinoma 9 4 T3N2M1 IV W43 80 Adenocarcinoma 7 4 T4NxM1 IV S44 80 Adenocarcinoma 5 ? T3N0M0 III S45 62 Adenocarcinoma 10 5 T3N2M1 IV S46 68 Adenocarcinoma 6 3 T3N0Mx W47 58 Adenocarcinoma 9 5 M1b IV S48 77 Adenocarcinoma 8 3 T2N0M0 II S49 65 Adenocarcinoma 9 ? T3N0M1 IV S50 66 Adenocarcinoma 9 5 TxN0M1 IV W51 86 Adenocarcinoma 9 3 T3NxM1b IV S52 48 Embryonal rhabdomyosarcoma 5 T4N0M0 IV N53 62 Adenocarcinoma 9 4 T4N0M0 IV S54 63 Adenocarcinoma 8 4 T4N0M0 IV S

(Continued)

Immunity to LEDGF/p75 in Prostate Cancer 21

proteolytic cleavage of specific proteins during tumorcell death could potentially generate, under inflamma-tory conditions, fragments with increased immuno-genicity that might trigger humoral and cell-mediatedimmune responses [28]. We demonstrated in a recentstudy that LEDGF/p75 is cleaved in apoptotic Jurkatand HeLa cells into fragments of 65 and 58 kDa, andthat this cleavage abrogates its pro-survival function[21]. It was, therefore, of interest to determine whetherLEDGF/p75 also undergoes cleavage during prostatetumor cell apoptosis. Absence of cleavage in prostatetumor cells would rule out the possibility that circulat-ing fragments of the protein derived from dying tumorcells could drive the production of autoantibodies toLEDGF/p75. Furthermore, absence of cleavage ofLEDGF/p75 in these dying cells could be alsoindicative of mutations in its caspase recognition sites,which would potentially enhance the pro-survivalfunction of the protein in tumors. To determine whe-ther LEDGF/p75 is cleaved during prostate tumor cellapoptosis, we induced apoptosis in LNCaP, PC3, andDU145 cells with either ActD/TRAIL or STS. Thesetreatments resulted in cleavage of LEDGF/p75 into itssignature caspase-induced fragments p65 and p58(Fig. 7), indicating that potentially immunostimulatoryproducts of the protein may be released in dyingprostate tumor cells. In DU145 and PC3 cells treatedwith ActD/TRAIL, only the 65 kDa fragment wasvisible. Since the 58 kDa fragment is a relatively latecleavage product of LEDGF/p75 [21], its absence fromthese blots is likely due to the slower rate of Act D/TRAIL-induced apoptosis in DU145 and PC3. Apopto-tic cleavage of LEDGF/p75 into these fragments wasnot specific for prostate tumor cells because it was alsoobserved in apoptotic, non-tumor PrEC cells (data notshown), consistent with our previous conclusion thatthis cleavage is a general feature of apoptosis [21].

DISCUSSION

There is compelling evidence that the production ofautoantibodies to intracellular proteins constitutes an

integral component of the anti-tumor immune re-sponse in cancer patients [1–7]. In this study, wescreened sera from 207 PCa patients, 166 age-matchedmen with no clinical evidence of PCa, 40 BPH patients,and 44 blood bank male donors for the presence ofautoantibodies to potential TAA. Using IIFmicroscopyand ELISA,we detected autoantibodies to LEDGF/p75in PCa patients at frequencies (14–18%) that weresignificantly higher than inmatched controls andbloodbank donors. These autoantibodies were not detectedin the BPH group. The frequency of these autoantibo-dies in the PCa group were within the range observedfor other autoantibodies in cancer [2,4]. For instance,Zhang et al. [2] detected by ELISA autoantibodies toseveral known TAA in our cohort of PCa patients atindividual frequencies ranging from 2.9 to 25.2%,depending on the antigen. This reinforced the notionthat the autoantibody response in human cancers ishighly heterogeneous and typically occurs at frequen-cies below 30% for individual antigens. However,increased detection of autoantibodies and definitionof distinct tumor-associated autoantibody profiles incancer can be achieved using uniquely constitutedTAA arrays [2,29].

We observed an incomplete correlation in thedetection of anti-LEDGF/p75 antibodies between thedifferent immunoassays (ELISA, immunoblotting, andIIF). This divergence was previously observed in otherautoantibody responses and could be attributed todifferences in the sensitivity of the various assays andin the antigen conformation in the individual assays[4,30]. Cancer-associated autoantibodies might displayheterogeneity in epitope recognition within a givenantigen. Consequently, some patients would produceautoantibodies against non-denatured epitopeswhereas others may produce antibodies against dena-tured epitopes. Thus, the combined use of differentassays may enhance detection of specific autoantibo-dies in a particular cancer type.

We hypothesized that the antibody response toLEDGF/p75 in PCa could be indicative of aberrantexpression of this protein in certain prostate tumors.

TABLE II. (Continued)

Sample Age Diagnosis GleasonField

Gleason TNM StageLEDGF/p75staining

55 59 Adenocarcinoma 9 ? M1 IV S56 59 Adenocarcinoma 9 ? T2N0M1 IV S57 48 Normal N58 75 Normal N59 71 Normal N

S, strong; W, weak; N, negative; U, unsatisfactory tissue; ?, could not be determined; *, pre-immune serum showed strong staining.

22 Daniels et al.

Because prostate tissue specimens from patients pro-ducing anti-LEDGF/p75 autoantibodies were notavailable in large numbers for this study, we examinedby immunohistochemistry the expression levels of theprotein in a commercially available prostate tissuearray. We observed that while the three normal pros-tate tissues samples in the array did not express orminimally expressed LEDGF/p75, 93% of the prostatetumor specimens expressed the protein. The majorityof these showed high expression (33 of 54, 61%).However, it was not possible to establish a correlationbetween strong immunostaining and tumor stage

because of the relatively small number of tumorsamples in stages I–III. The increased LEDGF/p75expression in prostate tumors and BPH compared tonormal prostate tissue suggested that this proteinmight be upregulated during prostate carcinogenesisor the development of inflammatory conditions of theprostate. This remains to be confirmed in a more com-plete analysis of LEDGF/p75 expression in normalprostate, prostatitis, BPH, prostatic inflammatoryatrophy (PIA), prostatic intraepithelial neoplasia (PIN),and prostate tumors at different stages.

LEDGF/p75 expression was also strong in morpho-logically normal epithelium located within the tumormicroenvironment. Chuaqui et al. [31] observed thatat the molecular level, normal epithelium adjacentto a prostate tumor might differ from epitheliumdistant from the tumor. These investigators noted thatadjacent normal tissue may be a tumor precursor, oralternatively, may be reactive to the invasive tumorwith an atypical genetic profile. This would be con-sistentwith the observation that PIA foci in the prostatehave low rates of apoptosis and high rates of cellproliferation and are often found adjacent to neoplasticlesions [32].

LEDGF/p75 is upregulated in normal cells inresponse to oxidative stress [18] and it is likely that itsrelatively high expression in prostate tumors and theiradjacent ‘‘normal’’ tissue is induced by inflammationand oxidative stress, key factors underlying prostatecarcinogenesis [33,34]. Ripple et al. [35] reported that anoxidative shift occurs during the aging process leadingto increased levels of oxidative stress in the prostate.There is also evidence for increased oxidative stress inthe benign epitheliumof PCa tissue compared to that ofnormal prostate tissue [36]. Inflammation and oxida-tive stress are also associated with development ofBPH [37], which could explain our observation thatLEDGFp75 is also strongly expressed in BPH tissue.The lack of autoantibodies to LEDGF/p75 in BPH seramight be indicative of relatively low rates of cell death,and therefore, decreased exposure of LEDGF/p75cleavage fragments, in BPH tissue.

LEDGF/p75 staining in prostate tumor and BPHtissue was predominantly confined to the cell nucleus,although cytoplasmic staining was also detected inmany cells. While it cannot be ruled out that thecytoplasmic staining results from non-specific bindingof the rabbit anti-LEDGF/p75 antibody, the nuclear/cytoplasmic localizationof theprotein inprostate tissuemay be indicative of its biological functions. LEDGF/p75 is known to be secreted and taken up by lensepithelial cells where it is transported through thecytosol into the nucleus, presumably via Tat-like andnuclear localization sequences [20,38,39]. LEDGF/p75 is a member of the hepatoma-derived growth

Fig. 6. ExpressionofLEDGF/p75 inPCa tissue.ImgenexHistoTM-ArrayPCa tissuesorindividualBPHtissueslideswerestainedwitharabbit anti-LEDGF/p75 antibody or the pre-immune serum, bothused at a1:500 dilution.A:Representativenormal tissues (HistoTM-Array specimens #58 and #59). B: Representative PCa tissues(HistoTM-Array specimens #53 and #24).C: Representative pic-ture of LEDGF/p75 expression in morphologically normal tissue(arrows) adjacent to the tumor area (HistoTM-Array specimen#22).D:RepresentativeBPHtissue.

Immunity to LEDGF/p75 in Prostate Cancer 23

factor family [39] and its exogenous addition to avariety of non-tumor cell lines stimulates proliferationand prolongs cell survival under stress conditions [20].LEDGF/p75 is also anuclear transport protein essentialfor the nuclear accumulation of HIV-1 integrase [40].

LEDGF/p75 protects cells against oxidative andheat stress by upregulating the levels of Hsps, partic-ularly Hsp27, Hsp90, ab-crystallin, and antioxidantprotein 2 (AOP2), via binding to stress responseelements and heat shock elements in Hsp gene

promoter regions [18,20,41]. Activation of Hsps resultsin an anti-apoptotic effect because these proteins inter-fere with caspases and other elements of the apoptoticmachinery [42–44]. Interestingly, Hsps, particularlyHsp27, have been implicated in prostate carcinogenesis[45–47]. It could be speculated that inflammation andincreased oxidative stress in the prostate tumormicroenvironment increase LEDGF/p75 expression,which in turn transcriptionally upregulates Hsps, re-sulting in a survival advantage to prostate tumor cells.

Increased expression of LEDGF/p75 in prostatetumors and caspase-mediated cleavage of the proteinin dying tumor cells might generate immunogenicforms of the protein which under the inflammatoryconditions of the tumor could elicit an autoantibodyresponse in certain PCa patients. We cannot rule out,however, that anti-LEDGF/p75 autoantibodies mightbe linked to an unrelated inflammatory disorder giventhat these autoantibodies have also been detected ininflammatory conditions suchas atopicdermatitis, age-related cataract, and asthma [48]. However, the signi-ficantly lower frequency of these autoantibodies in thematched control group would argue against thispossibility. Moreover, analysis of available clinicalrecords of PCa patients positive for anti-LEDGF/p75antibody failed to reveal previous history of any ofthese conditions for most patients. It should be notedthat inflammation, oxidative stress, and aberrant celldeath have also been implicated in the pathogenesis ofatopicdermatitis, age-related cataract, andasthma [49–54]. This raises the possibility that humoral immunityto LEDGF/p75 could be a feature of certain diseaseconditions, including PCa, associated with these threefactors. Although anti-LEDGF/p75 antibodies were notdetected in our BPH group, we cannot rule out thatother common prostate inflammatory conditions, suchas prostatitis, may give rise to these autoantibodies.

In summary, our data show that certain patientswith PCa produce autoantibodies to the survivalprotein LEDGF/p75 and that this protein is highly

TABLE III. Summaryof Data on LEDGF/p75 Staining of ProstateTumorTissues on the ImgenexHisto-TM-Array in RelationtoClinical Stage

Positive LEDGF/p75staining

Strongstaining

Weakstaining

Negative or minimalstaining

Normal prostate (n¼ 3) 0 (0%) 0 (0%) 0 (0%) 3 (100%)PCa tumor number (n¼ 54) 50 (93%) 33 (61%) 17 (31%) 4 (6%)Clinical stage I (n¼ 1) 1 (100%) 0 (0%) 1 (100%) 0 (0%)Clinical stage II (n¼ 12) 11 (92%) 5 (42%) 6 (50%) 1 (8%)Clinical stage III (n¼ 3) 3 (100%) 3 (100%) 0 (0%) 0 (0%)Clinical stage IV (n¼ 35) 32 (91%) 23 (66%) 9 (26%) 3 (9%)

LEDGF/p75, lens epithelium-derived growth factor p75.

Fig. 7. Cleavage of LEDGF/p75 in PCa cell lines undergoing apop-tosis.A: Representative immunoblot blot showing the cleavage ofLEDGF/p75 into 65 and 58 kDa fragments in Jurkat T cells (used aspositive control) andDU145 cells induced to dieby apoptosiswith 2and 4 mM STS, respectively.B: Apoptosis was induced in LNCaP,DU145, and PC3 cells by exposure to100 ng/ml Act D for 4 hr fol-lowedby treatmentwith100ng/mlTRAILforupto24hr.Bandsweredetectedin theblotsusing serumfromPCapatientP096.

24 Daniels et al.

expressed in prostate tumor tissue and cleaved intodistinct fragments in apoptotic prostate cells. If demo-nstrated to beupregulated in the prostate in response toincreased oxidative stress and to play a pro-survivalfunction during prostate carcinogenesis, LEDGF/p75could become a target of novel therapeutic strategiesaimedatmodulating itspro-survival functionorknock-ing down its expression in the malignant prostate.

ACKNOWLEDGMENTS

We are grateful to Dr. Eng M. Tan (The Scripps Re-search Institute, La Jolla, CA), Dr. Michael Lilly (LomaLinda University School of Medicine), Dr. DanielNachtsheim (Scripps Clinic, La Jolla, CA), andDr. Ann Rearden (University of California San Diego,La Jolla, CA) for helpful discussions and technicaladvice. We also thank Dr. Tan for kindly providing theNHS used in this study. We are grateful to Dr. HerbertRuckle (Loma Linda University School of Medicine,Loma Linda, CA) for facilitating the acquisition of serafromPCapatients and toDr. KerbyOberg andShannonThorn (Loma Linda University School of Medicine,Loma Linda, CA) for technical assistance in the use ofthe Biogenic i6000 auto stainer. Dr. Floyd Petersen(Loma Linda University School of Public Health, LomaLinda, CA) kindly provided statistical consultation.

REFERENCES

1. ScanlanMJ, Welt S, Gordon CM, Chen YT, Gure AO, Stockert E,Jungbluth AA, Ritter G, Jager D, Jager E, Knuth A, Old LJ.Cancer-related serological recognition of human colon cancer:Identification of potential diagnostic and immunotherapeutictargets. Cancer Res 2002;62:4041–4047.

2. Zhang JY, Casiano CA, Peng XX, Koziol JA, Chan EK, Tan EM.Enhancement of antibody detection in cancer using panel ofrecombinant tumor-associated antigens. Cancer EpidemiolBiomarkers Prev 2003;12:136–143.

3. Stockert E, Jager E, Chen YT, Scanlan MJ, Gout I, Karbach J,Arand M, Knuth A, Old LJ. A survey of the humoral immuneresponse of cancer patients to a panel of human tumor antigens.J Exp Med 1998;187:1349–1354.

4. Rohayem J, Diestelkoetter P, Weigle B, Oehmichen A, SchmitzM, Mehlhorn J, Conrad K, Rieber EP. Antibody response to thetumor-associated inhibitor of apoptosis protein survivin incancer patients. Cancer Res 2000;60:1815–1817.

5. Tan EM. Autoantibodies as reporters identifying aberrantcellular mechanisms in tumorigenesis. J Clin Invest 2001;108:1411–1415.

6. Tan EM, Shi FD. Relative paradigms between autoantibodies inlupus and autoantibodies in cancer. Clin Exp Immunol 2003;134:169–177.

7. Soussi T. p53 Antibodies in the sera of patients with varioustypes of cancer: A review. Cancer Res 2000;60:1777–1788.

8. Ablin RJ. Malignancy associated with antinuclear antibodies.Lancet 1972;2:1253–1254.

9. Ablin RJ. Serum antibody in patients with prostatic cancer. Br JUrol 1976;48:355–361.

10. Liang T, Hiipakka RA, Stebbins J, Liao S. Anti-5 alpha-reductaseautoantibodies in the serum of patients with prostatic cancer. JClin Endocrinol Metab 1990;71:1666–1668.

11. Lang C, Unteregger G, Kartarius S, Gunther J, Bonkhoff H,Montenarh M, Zwergel T. p53 autoantibodies in patients withurological tumours. Br J Urol 1998;82:721–726.

12. Nilsson BO, Carlsson L, Larsson A, Ronquist G. Autoantibodiestoprostasomes asnewmarkers forprostate cancer.Ups JMedSci2001;106:43–49.

13. Mintz PJ, Kim J, Do K-A, Wang X, Zinner RG, Cristofanilli M,Arap MA, Hong WK, Troncoso P, Logothetis CJ, Pasqualini R,Arap W. Fingerprinting the circulating repertoire of antibodiesfrom cancer patients. Nat Biotechnol 2003;21:57–63.

14. Mollick JA, Hodi FS, Soiffer RJ, Nadler LM, Dranoff G. MUC1-like tandem repeat proteins are broadly immunogenic in cancerpatients. Cancer Immunol 2003;3:3.

15. Zhou Y, TothM,HammanMS,Monahan SJ, Lodge PA, BoyntonAL, Salgaller ML. Serological cloning of PARIS-1: A new TBCdomain-containing, immunogenic tumor antigen from a pros-tate cancer cell line. Biochem Biophys Res Commun 2002;290:830–838.

16. Fossa A, Alsoe L, Crameri R, Funderud S, Gaudernack G,Smeland EB. Serological cloning of cancer/testis antigensexpressed in prostate cancer using cDNAphage surface display.Cancer Immunol Immunother 2004;53:431–438.

17. Ge H, Si Y, Roeder RG. Isolation of cDNAs encoding noveltranscription coactivators p52 and p75 reveals an alternateregulatory mechanism of transcriptional activation. EMBO J1998;17:6723–6729.

18. Singh DP, Ohguro N, Chylack LT, Jr, Shinohara T. Lensepithelium-derived growth factor: Increased resistance tothermal and oxidative stresses. Ophthalmol Vis Sci 1999;40:1444–1451.

19. Ochs RL,Muro Y, Si Y, GeH, Chan EK, Tan EM. Autoantibodiesto DFS 70 kDa/transcription coactivator p75 in atopic dermatitisand other conditions. J Allergy Clin Immunol 2000;105:1211–1220.

20. Shinohara T, Singh DP, Fatma N. LEDGF: A survival factor,activates stress-related genes. Prog Retinal Eye Res 2002;21:341–358.

21. Wu X, Daniels T, Molinaro C, Lilly MB, Casiano CA. Caspasecleavage of the nuclear autoantigen LEDGF/p75 abrogates itspro-survival function: Implications for autoimmunity in atopicdisorders. Cell Death Differ 2002;9:915–925.

22. Mikuz G. Pathology of prostate cancer. Old problems and newfacts. Adv Clin Pathol 1997;1:21–34.

23. Zhang JY, ChanEK, PengXX, LuM,WangX,Mueller F, TanEM.Autoimmune responses tomRNAbinding proteins p62 andKocin diverse malignancies. Clin Immunol 2001;100:149–156.

24. Wu X, Molinaro C, Johnson N, Casiano CA. Secondary necrosisis a source of proteolytically modified forms of specificintracellular autoantigens. Arthritis Rheum 2001;44:2642–2652.

25. Casiano CA, Ochs RL, Tan EM. Distinct cleavage products ofnuclear proteins in apoptosis and necrosis revealed by autoanti-body probes. Cell Death Differ 1998;5:183–190.

26. Utz PJ, Gensler TJ, Anderson P. Death, autoantigen modifica-tions, and tolerance. Arthritis Res 2000;2:101–114.

27. Rosen A, Casciola-Rosen L. Autoantigens as substrates forapoptotic proteases: Implications for the pathogenesis ofsystemic autoimmune disease. Cell Death Differ 1999;6:6–12.

28. Savill J, Fadok V. Corpse clearance defines the meaning of celldeath. Nature 2000;407:784–788.

Immunity to LEDGF/p75 in Prostate Cancer 25

29. von Brevern MC, Hollstein MC, Cawley HM, De Benedetti VM,Bennet WP, Liang L, He AG, Zhu SM, Tursz T, Janin N, TriversGE. Circulating anti-p53 antibodies in esophageal cancerpatients are found predominantly in individuals with p53core domain mutations in their tumors. Cancer Res 1996;56:4917–4921.

30. Koziol JA, Zhang JY, Casiano CA, Peng XX, Shi FD, Feng AC,Chan EK, Tan EM. Recursive partitioning as an approach toselection of immune markers for tumor diagnosis. Clin CancerRes 2003;9:5120–5126.

31. Chuaqui RF, Englert CR, Strup SE, Vocke CD, Zhuang Z, DurayPH, Bostwick DG, Linehan WM, Liotta LA, Emmert-Buck MR.Identification of a novel transcript up-regulated in a clinicallyaggressive prostate carcinoma. Urology 1997;50:302–307.

32. De Marzo AM, Marchi VL, Epstein JI, Nelson WG. Proliferativeinflammatory atrophy of the prostate: Implications for prostaticcarcinogenesis. Am J Pathol 1999;155:1985–1992.

33. NelsonWG, DeMarzo AM, IsaacsWB. Prostate cancer. N Engl JMed 2003;349:366–381.

34. Lucia MS, Torkko KC. Inflammation as a target for prostatecancer chemoprevention: Pathological and laboratory rationale.J Urol 2004;171:S30–S34.

35. Ripple MO, Hentry WF, Rago RP, Wilding G. Prooxidant-antioxidant shift induced by androgen treatment of humanprostate carcinoma cells. J Nat Cancer Inst 1997;89:40–48.

36. Liu Z, Wang LE, Strom SS, Spitz MR, Babaian RJ, DiGiovanni J,WeiQ.Overexpression of hMTH in peripheral lymphocytes andrisk of prostate cancer: A case-control analysis.Mol Carcinogen-esis 2003;36:123–129.

37. De Marzo AM, Coffey DS, Nelson WG. New concepts in tissuespecificity for prostate cancer and benign prostatic hyperplasia.Urology 1999;53:29–39.

38. FatmaN, SinghDP, Shinohara T, Chylack LT, Jr. Heparin’s rolesin stabilizing, potentiating, and transporting LEDGF into thenucleus. Invest Ophthalmol Vis Sci 2000;41:2648–2657.

39. Singh DP, Ohguro N, Kikuchi T, Sueno T, Reddy VN, Yuge K,Chylack LT, Jr, Shinohara T. Lens epithelium-derived growthfactor: Effects on growth and survival of lens epithelial cells,keratinocytes, and fibroblasts. Biochem Biophys Res Commun2000;267:373–381.

40. MaertensG,CherepanovP, PluymersW,BusschotsK,DeClercqE,DebyserZ, EngelborghsY. LEDGF/p75 is essential for nuclearand chromosomal targeting of HIV-1 integrase in human cells.J Biol Chem 2003;278:33528–33539.

41. FatmaN, SinghDP, Shinohara T, Chylack LT, Jr. Transcriptionalregulation of the AOP2 gene: A thiol-specific antioxidant, by

LEDGF to protect cells from oxidative stress. J Biol Chem 2001;276:48899–48907.

42. Bruey JM, Ducasse C, Bonniaud P, Ravagnan L, Susin SA,Diaz-Latoud C, Gurbuxani S, Arrigo AP, Kroemer G, Solary E,Garrido C. Hsp27 negatively regulates cell death by interactingwith cytochrome c. Nat Cell Biol 2000;2:645–652.

43. Kamradt MC, Chen F, Cryns VL. The small heat shockprotein alpha B-crystallin negatively regulates cytochromec- and caspase-8-dependent activation of caspase-3 by inhibit-ing its autoproteolytic maturation. J Biol Chem 2001;276:16059–16063.

44. Parcellier A, Gurbuxani S, Schmitt E, Solary E, Garrido C. Heatshockproteins, cellular chaperones thatmodulatemitochondrialcell death pathways. Biochem Biophys Res Commun 2003;304:505–512.

45. Cornford PA, Dodson AR, Parsons KF, Desmond AD, Woolfen-denA, FordhamM,Neoptolemos JP,KeY, FosterCS.Heat shockprotein expression independently predicts clinical outcome inprostate cancer. Cancer Res 2000;60:7099–7105.

46. Lebret T,Watson RW, Fitzpatrick JM.Heat shock proteins: Theirrole in urological tumors. J Urol 2000;169:338–346.

47. Gibbons NB,Watson RW, Coffey RN, BradyHP, Fitzpatrick JM.Heat-shock proteins inhibit induction of prostate cancer cellapoptosis. Prostate 2000;45:58–65.

48. Ganapathy V, Daniels T, Casiano CA. LEDGF/p75: A novelnuclear autoantigen at the crossroads of cell survival andapoptosis. Autoimmunity Rev 2003;2:290–297.

49. Fuchs J, Zollner TM, Kaufmann R, Podda M. Redox-modulatedpathways in inflammatory skin diseases. Free Radic Biol Med2001;30:337–353.

50. Bowler RP, Crapo JD. Oxidative stress in allergic respiratorydiseases. J Allergy Clin Immunol 2002;110:349–356.

51. Ottonello S, Foroni C, Carta A, Petrucco S, Maraini G. Oxidativestress and age-related cataract. Ophthalmologica 2000;214:78–85.

52. Trautmann A, Akdis M, Klunker S, Blaser K, Akdis CA. 2001.Role of apoptosis in atopic dermatitis. IntArchAllergy Immunol124:230–232.

53. Trautmann A, Schmid-Grendelmeier P, Kruger K, Crameri R,Akdis M, Akkaya A, Brocker EB, Blaser K, Akdis CA. T cellsand eosinophils cooperate in the induction of bronchial epi-thelial cell apoptosis in asthma. J Allergy Clin Immunol 2002;109:329–337.

54. Shinohara T, Singh DP, Chylack LT, Jr. Review: Age-relatedcataract: Immunity and lens epithelium-derived growth factor(LEDGF). J Ocul Pharmacol Ther 2000;16:181–191.

26 Daniels et al.