The Prostate 60:153 ^159 (2004)

Interleukin-8 Expression Is Increasedin SenescentProstatic Epithelial Cells and Promotes the

Developmentof Benign ProstaticHyperplasia

Patricia Castro,1,4 Chen Xia,1,4 Lori Gomez,1,4 Dolores J. Lamb,2,3

and Michael Ittmann1,4*1Departmentof Pathology, Baylor College ofMedicine,One Baylor Plaza,Houston,Texas

2ScottDepartmentofUrology,Baylor College ofMedicine,One Baylor Plaza,Houston,Texas3MolecularandCell Biology, Baylor College ofMedicine,One Baylor Plaza,Houston,Texas

4HoustonDepartmentof Veterans AffairsMedical Center, 2002Holcombe Blvd,Houston,Texas

BACKGROUND. Benignprostatichyperplasia (BPH) is anextremely commondiseaseof oldermen characterized by increased growth of prostatic epithelial and stromal cells. Previously weshowed that senescent epithelial cells accumulate in the prostate of aging men and secreteinterleukin-1a (IL-1a). IL-8 is also present at increased levels in BPH tissues and inducesexpression of FGF2, a potent stromal growth factor. Therefore, we sought to determine if IL-8 isalso expressed at increased levels by senescent epithelial cells and if this secreted IL-8 plays arole in the pathogenesis of BPH.METHODS. Expression of IL-8 in human BPH tissue and primary cultures of prostaticepithelial cellswas analyzedusing an enzyme-linked immunoabsorption assay (ELISA). Tissuesenescencewasassessedbyaquantitative assay for senescence-associatedbgalactosidase (SA-bgal). Proliferation of primary and immortalized prostatic epithelial cells in response to IL-8wasdetermined by counting of cells at intervals after addition of IL-8.RESULTS. Expression of IL-8 is significantly increased in vitro when cultured prostaticepithelial cells undergo senescence. Quantitative assay of BPH tissue extracts revealed thattissue IL-8 levels are correlated with both SA-b gal activity and prostate weight. IL-8 promotesproliferation of primary and immortalized prostfatic epithelial cells in culture.CONCLUSIONS. Senescence of prostatic epithelial cells results in increased expression ofIL-8,which canpromoteproliferation of non-senescent epithelial and stromal cells bydirect andindirectmechanisms, and in thismanner contributes to the increased tissuegrowth seen inBPH.Prostate 60: 153–159, 2004. # 2004 Wiley-Liss, Inc.

KEY WORDS: benign prostatic hyperplasia; cytokines; senescence

INTRODUCTION

Cellular senescence is a process that limits the pro-liferation of human cells [1,2]. The senescence responsecan be induced by a variety of cellular alterationssuch as loss of telomeric DNA, inappropriate ex-pression of oncogenes, DNA damage, and oxidativestress [1–5]. Senescent cells accumulate in tissues [1](including the prostate [6,7])with increasing age. Thesesenescent cells have altered function, including in-creased expression of cytokines and proteases anddecreased expression of protease inhibitors, all ofwhich may alter the function of adjacent cells [7–12].

Thus, the senescent cells that accumulate with increas-ing agemay contribute to the agingphenotype and age-related pathologies by secreting factors that act in a

Grant sponsor: NIDDK; Grant number: R01 DK54170; Grantsponsor: National Institute of Health (to D.J.L.); Grant number:T32 DK07763.

*Correspondence to: Michael Ittmann, MD, PhD, Research Service,Houston VAMC, 2002 Holcombe Blvd, Houston, TX 77030.E-mail: mittmann@bcm.tmc.eduReceived 21 October 2003; Accepted 10 December 2003DOI 10.1002/pros.20051Published online 2 February 2004 in Wiley InterScience(www.interscience.wiley.com).

� 2004 Wiley-Liss, Inc.

paracrine manner on adjacent cells and extracellularmatrix.

Benign prostatic hyperplasia (BPH) is an extremelycommon disease of older men, occurring in more than70%ofmen over the age of 60, and results in substantialmorbidity in this patientpopulation [13].Amajor factorin the pathogenesis of prostatic hyperplasia is the con-tinuing growth of the transition zone of the prostatedue to both epithelial and stromal proliferation. Wereported that senescent prostatic epithelial cells accu-mulate within the prostate with increasing age andthese cells secrete interleukin-1a (IL-1a) [7]. Consistentwith this observation, tissue IL-1a levels are increasedin BPH tissue [14]. IL-1a can induce expression of FGF7by prostatic stromal cells and tissue IL-1a levels arestrongly correlated with FGF7 levels, a potent prostaticepithelial growth factor. Tissue FGF7 levels are sub-stantially increased in BPH and strongly correlatedwith prostatic epithelial proliferation in vivo [15]. Thus,senescence of prostatic epithelial cells drives prolifera-tion of non-senescent epithelial cells via a dual para-crine mechanism involving secretion of FGF7 byadjacent stromal cells in response to IL-1a secreted bysenescent epithelial cells.

Increased proliferation of stromal cells also occursin BPH. We demonstrated that FGF2 (basic FGF) ismarkedly increased in BPH and increased levels ofFGF2 are correlated with increased stromal prolifera-tion [15]. IL-8 is secreted by primary epithelial cells inculture and induces FGF2 expression by prostaticstromal cells [16]. Based on our observation that IL-1asecretion is induced by senescence of prostatic epi-thelial cells, we sought to determine whether IL-8 ex-pression is induced in a similarmanner.Wehave foundthat the expression of IL-8 is significantly increasedin vitro when cultured prostatic epithelial cells under-go senescence. By quantitative assay of BPH tissueextracts, tissue IL-8 levels are correlated with bothsenescence-associated b galactosidase (SA-b gal) activ-ity and prostate weight. We demonstrate that inaddition to its ability to induce FGF2 expression instromal cells, IL-8 can directly promote proliferation ofprimary and immortalized prostatic epithelial cells inculture. Thus, senescence of prostatic epithelial cellsinduces expression of IL-8, which in turn can promoteproliferation of non-senescent epithelial cells directlyand of adjacent stromal cells by induction of FGF2.

MATERIALSANDMETHODS

TissueAcquisition

Samples of the benign tissue from hyperplastic tran-sition zone were taken from radical prostatectomiesand snap frozen in liquid nitrogen. The tissues werefree of carcinoma, determined as described previously

[17]. All tissues were collected with the approval of theBaylor College of Medicine Institutional Review Boardfor Human Studies.

Cell Culture

Immortalized prostatic epithelial cells (PNT1A)were maintained in RPMI 1640 with 10% fetal bovineserum. Primary cultures of prostatic epithelial cellswere grown in complete growth medium followingcollagenase digestion as described previously [15]. Thegrowth medium consisted of MCDB 153/RPMI 1640(9:1) containing epidermal growth factor (10 ng/ml),bovine pituitary extract (100 mg/ml), insulin, transfer-rin, and selenium (1% ITS; Sigma Chemicals, St. Louis,MO), dexamethasone (1 mM), bovine serum albumin(100 g/ml) with oleic acid (0.4 mg/ml), heparin (50 mg/ml), cholera toxin (0.1 mg/ml), 1% fetal bovine serum,and gentamycin (50 mg/ml). For determination of IL-8expression, subconfluent epithelial cells were split andplated at 5� 104 cells per 35mmdish.Cell extractswereprepared as described previously [16] on days 1, 2, 4,and 6 after plating. For cell growth assays, 5� 104

primary epithelial cells were plated on each collagen-coated 35 mm dish in complete epithelial growthmedium without 1% serum. Cells were kept in thismedium as controls or were supplemented with 0.1,1.0, or 10.0 ng/ml of recombinant IL-8 (R&D Systems,Minneapolis, MN). Cells were then trypsinized andcollected after 24 or 72 hr and counted using a Coultercounter. Similar experiments were performed withPNT1A cells using uncoated 35 mm dishes with RPMI1640 media supplemented with 1% ITS.

Enzyme-Linked ImmunoabsorptionAssay (ELISA)

Cell extractswere prepared from snap frozen tissuesor cell extracts as described previously [16] and pro-tein concentration determined using BioRad ProteinAssay following manufacturer’s instructions (BioRad,Hercules, CA). Determination of the IL-8 and IL-1aconcentration in cell and tissue extracts was carried outby ELISA based on a quantitative sandwich immuno-assay technique as described previously [14,16].

QuantitativeAssessmentof SA-b gal Activity

Tissue lysates prepared as described above wereused for quantitative SA-b gal assay as describedpreviously [7]. Briefly, 50mg of total proteinwasdilutedwith lysis buffer to a total volume of 50 ml and placed induplicate microtiter wells and 100 ml of SA-b gal stainsolution [18], pH 6.0, was added to each well. The platewas then incubated at 378C for 20 hr and theOD read at590 nm.One unit of SA-b gal activitywas defined as the

154 Castro et al.

activity leading to one OD unit at 590 nm in a 20-hrincubation.

Determination of the Prostate Specif|c Antigen(PSA)Contentof Extracts

PSAwas measured in diluted tissue extracts using asolid phase-two site immunoradiometric assay fromHybritech, Inc. (San Diego, CA) as described pre-viously [15]. Correction using PSA content is based onthe assumption that PSA per luminal epithelial cell isrelatively constant.

Statistical Analysis

Data was analyzed for statistical significance usingSigmaStat SoftwareTM (SPSS, Inc., Chicago, IL). Statis-tical significance was determined by ANOVA orStudent’s t-test with statistical significance defined asP< 0.05.

RESULTS

IL-8 Secretion isMarkedly Increasedin SenescentProstatic Epithelial Cells

To determine if senescent prostatic epithelial cellssecrete more IL-8, primary epithelial cultures wereestablished from prostate tissue after collagenasedigestion. These cultures were maintained until theywere subconfluent, cells split andplated. Previouslyweshowed that under these culture conditions, the epi-thelial cells became senescent after 6 days in culture asdetermined by the induction of p16 protein (expressionof SA-b gal), and the appearance of typical senescentmorphology. Therefore, cellswere collected after 1, 2, 4,or 6 days in culture and the IL-8 content determined byELISA.As shown in Figure 1, there is amarked increasein IL-8 content (per mgprotein) after 6 days in culture, atwhich time a large number of cells showed a typicalsenescent morphology. Thus, IL-8 expression is in-creased as epithelial becomes senescent in vitro.

Todetermine if IL-8 expression is correlatedwith thepresence of senescent epithelial cells in vivo, we quan-titatively determined the SA-b gal activity and IL-8content in protein extracts of the prostatic transitionzone collected from men undergoing radical prosta-tectomy. All of these tissues were characterized bythe presence of mild to severe BPH and were free ofcarcinoma. We previously showed that the assay forSA-bgal activity is pHdependent,with optimal activityat pH 6.0 and is correlated to the extent of SA-b galstaining determined histochemically on frozen sectionsof the same tissues. Both SA-b gal and IL-8 are expres-sed primarily by the prostatic epithelial cells [7,16] so asto correct for the variable percentage of epithelium in

tissues used for each extract, the level of PSA in eachextract was measured as described previously [15] andthe SA-b gal and IL-8 content per mg PSA calculated.The tissue samples were then divided into threeroughly equal groups: low SA-b gal (<1.0 U SA-bgal/mg PSA), intermediate (1–3 U/mg PSA), and high(>3 U/mg PSA) and the IL-8 content in pg per mg PSAcalculated for each group.

As shown in Figure 2, there is a very strongcorrelation between the tissue SA-b gal activity andIL-8 content after correcting for variability in epithelialcontent of the tissues that is statistically significant(P< 0.01, ANOVA). Consistent with this observation,there is also a clear trend for tissue IL-8 content toincrease with age (Fig. 3), although this correlation isnot statistically significant. We have previously shownthat the tissue content of IL-1a is correlated with SA-bgal activity. As shown in Figure 4, there is a very goodcorrelation of tissue IL-8 content and tissue IL-1a con-tent, although this correlation only approaches statis-tical significance (P¼ 0.056). Taken together, this dataindicates that tissue senescence is significantly corre-lated with tissue IL-8 levels in BPH.

IL-8 secreted by senescent prostatic epithelial cellscan induce the expression of FGF2 in adjacent stromalcells, which in turn can act as an autocrine growthfactor for these stromal cells, and to a lesser extent, asa paracrine growth factor for non-senescent prostaticepithelial cells [16]. To determine if IL-8 can also actdirectly as an epithelial growth factor, we treated pri-mary prostatic epithelial cells with IL-8 (0.1–10 ng/ml)

Fig. 1. Interleukin-8 (IL-8) contentofprimaryprostaticepithelialcells undergoing senescenceprimarycultures of prostatic epithelialcells were established after collagenase digestion andreplated.Celllysates were collected after 1, 2, 4, and 6 days in culture and IL-8content determined by enzyme-linked immunoabsorption assay(ELISA). Cells had a senescent morphology by day 6. Mean� therangeofduplicatevaluesis shown.

Interleukin-8 in the Pathogenesis of BPH 155

in serum-free medium. As can be seen in Figure 5A,IL-8 is a growth factor for primary epithelial cells, witha statistically significant increase in cell number(P< 0.05, t-test) for cells treated with either 1 or 10 ng/

ml of IL-8 for 72 hr relative to control cells incubatedwithout IL-8. Similar results were seen using theimmortalized prostatic epithelial cell line PNT1A(Fig. 5B), with a statistically significant increase in cellnumber at 72 hr for cells treated with 10 ng/ml IL-8relative to control cells (P< 0.05, t-test). These concen-trations of IL-8 are biologically relevant since BPHtissue contains approximately 33 ng IL-8 per gram oftissue on an average. Thus IL-8 can promote prostaticgrowth, both as a direct growth factor for prostaticepithelial cells and as a paracrine inducer of FGF2.

Given that tissue IL-8 levels are increased in BPHtissue, and IL-8 can promote proliferation of prostaticepithelial and stromal cells, we examined the correla-tion between tissue IL-8 levels and prostate weight.Prostateswere divided into three groups on the basis ofprostate weight: mild (<40 g), moderate (40–60 g), andsevere (>60 g) BPH and the mean tissue IL-8 contentdetermined. As shown in Figure 6, there is a strongcorrelation between tissue IL-8 levels and prostateweight. This correlation is statistically significant byANOVA (P< 0.05). Thus tissue IL-8 levels and byimplication, epithelial senescence, have amajor impacton the pathogenesis of BPH.

DISCUSSION

We report here that senescent prostatic epithelialcells secrete IL-8 in vitro. There is a strong correlationbetween tissue senescence and levels of IL-8 in theprostate and further that IL-8 levels are significantlycorrelated with prostate weight. Previously we report-

Fig. 2. CorrelationofIL-8andsenescence-associatedbgalactosi-dase (SA-b gal) activity in prostate extracts. SA-b gal activity wasdetermined in prostate extracts and IL-8 content quantitated onthe same extracts by ELISA.The content of SA-b gal activity andIL-8 ineachextractwascorrectedforvariationinepithelialcontentusingthecontentofprostatespecificantigen(PSA),anepithelialspe-cific protein, determined on each extract by radioimmunoassay.Extracts were divided into three roughly equal size groups basedon SA-b gal activity per mg/PSA: Low<1.0; moderate1.0 ^3.0, andhigh>3.0.The mean and standard error of the IL-8 content (in ngpermgPSA)of theextractsineachgroupisindicated.Thedifferencesbetween groups were statistically significant (P< 0.008, ANOVA).[Color figure canbeviewedin the online issue,which is available atwww.interscience.wiley.com.]

Fig. 3. Relationship of IL-8 content in prostate extracts andpatient age.The IL-8 content of benign prostatic hyperplasia (BPH)tissue extracts was determined by ELISA. Patients are grouped byage in decades.Themean and the standarderror of the IL-8 contentpermgPSAareshownforeachagegroup.[Color figurecanbeviewedintheonlineissue,whichisavailableatwww.interscience.wiley.com.]

Fig. 4. Correlation of tissue IL-8 and IL-1a content The con-centration of IL-1a and IL-8 protein in each BPH tissue extractwas determined by ELISA. Tissues were grouped into threegroups as shown based on tissue levels of IL-8: 4 pg/mg protein(n¼ 5), 4 ^14 pg/mg protein (n¼ 8), and >14 pg/mg protein (n¼ 9).The mean� the standard error for the IL-1a content per mgprotein in each group is shown. [Color figure can be viewed in theonlineissue,whichis availableatwww.interscience.wiley.com.]

156 Castro et al.

ed that IL-8 induces FGF2 expression by adjacentprostatic stromal cells via increased expression of FGF2mRNA. This FGF2 can act as a potent autocrine stromalgrowth factor and, to a lesser extent as a paracrineepithelial growth factor. We report here that IL-8 canalso directly promote growth of both benign prostaticepithelial cells in culture. Accordingly, IL-8 can bothdirectly and indirectly promote proliferation of pro-static epithelial cells. IL-8 is also a well-known pro-moter of angiogenesis [19] and this activity may also beimportant in sustaining the increased mass of theprostatic transition zone in BPH.

We have demonstrated previously that IL-1a caninduce expression of FGF7 by prostatic stromal cellsand tissue IL-1a levels are strongly correlated withFGF7 levels [14]. Tissue FGF7 levels are in turn strongly

correlatedwithprostatic epithelial proliferation invivo.This mechanism would enhance the direct effects ofIL-8 on epithelial proliferation described above.Of note, the observed correlation between IL-8 levelsand prostate weight may reflect the combined actionsof these two cytokines since both are co-regulatedduring epithelial senescence. Consequently, senes-cence of prostatic epithelial cells drives proliferationof non-senescent epithelial cells and stromal cells bydirect and indirect actions of secreted cytokines. Ourcurrent results provide an additional mechanistic linkbetween alterations induced by cellular senescence andthe pathogenesis of a disease that is strongly associatedwith age.

Prostatic epithelial cellular senescence may contri-bute to prostatic hyperplasia by other mechanisms inaddition to secretion of IL-8 and IL-1a. Senescent cellsare often resistant to apoptosis [20], so accumulation ofsenescent cells may directly increase the number ofepithelial cells in BPH tissue. Other investigators haveshown that cytokines, growth factors, andproteases aresecreted by senescent prostatic epithelial cells in vitro.For example, an analysis of gene expression by serialanalysis of gene expression (SAGE) revealed thatparathyroid-hormone related protein (PTHRP), cathe-psin B, and MMP-14 are all expressed at increasedlevels by senescent prostatic epithelial cells while TGF-B1, a potent epithelial growth repressor, was decreased[11]. A similar study using cDNA microarray analysis[12] revealed increased expression of MMP-2, cathe-psin F, carboxypeptidase M, and insulin-like growthfactor 2. Proteases can potentially increase proliferation

Fig. 5. Proliferation of prostatic epithelial cells treated withrecombinant IL-8.Cells were plated and treated IL-8 at 0.1, 1.0, or10ng/mlandcountedattheindicatedtimes.A:Primaryepithelialcellcultures. B: PNT1A, an immortalized prostatic epithelial cell line.Error bars indicate the standard deviation of triplicate determina-tions.The statistical significance of the difference between the cellnumber of each treatment group and controls was determined byStudent’s t-test and thosegroups forwhichP< 0.05 areindicatedbyanasterisk.

Fig. 6. Correlation of tissue IL-8 content and prostate weight.Tissue concentration of IL-8 was determined in tissue extracts byELISA.Themeanandstandarderrorof the tissueIL-8concentrationfor prostates with mild (prostate weight<40 g, n¼ 4), moderate(40 ^ 60 g, n¼ 7), or severe (>60 g, n¼11) BPH are shown. [Colorfigure can be viewed in the online issue, which is available atwww.interscience.wiley.com.]

Interleukin-8 in the Pathogenesis of BPH 157

by releasing bound growth factors from extracellularmatrix and/or specific binding proteins. PTHRP isknown to be expressed in prostate cancer and can act asa growth factor for prostate cancer cells [21], but basedon immunohistochemical studies it is also expressed ina subgroup of epithelial cells in BPH tissue [22]. IGF-2 isalso expressed focally in prostatic epithelial cells [23]and is present in substantial quantities in BPH tissue(121 ng/g tissue) [24]. Focal staining of PTHRP andIGF-2 in the epithelial cells is similar to what Castroet al. [7] and Choi et al. [6], have observed for SA-b galby histochemistry and contrasts to the more diffusestaining seen with other markers such as PSA orkeratin-18. Of note is the finding that IGF-2 is increased8.3 fold in the prostate of aging Wistar rats incomparison to young adult controls [25]. Furtherstudies are needed to confirm the in vivo expressionof these proteins and their role in promoting prostaticgrowth in BPH.

The presence of significant numbers of senescentepithelial cells in the prostate and the lack of senescentstromal cells, as observed by us and others, is in con-trast to the findings in other organs such as skin, wheresenescent stromal cells are more prominent [18]. Thismay reflect the low turnover rate of prostatic epithe-lium, as assessed by the relatively low number of pro-liferating cells [15], in contrast to other epithelia such asskin, that have a much higher turnover rate. This lowmitotic activity also makes it less likely that there issignificant loss of telomeres within the prostatic epi-thelium. It is thus likely that othermechanisms, such asoxidative stress and/orDNAdamage, are inducing thesenescence response in the prostatic epithelial cells. Inthis regard it should be noted that while there is ageneral correlation of increased IL-1a and IL-8 withincreasing age, this correlation is not statistically signi-ficant, while the correlation of SA-b gal activity withlevels of both of these cytokines is statisticallysignificant. This implies that epithelial senescence,and not age per se, is driving increased expression ofthese cytokines and that there are probably indivi-dual patient factors (environmental and/or genetic),in addition to increasing age that promote epithelialsenescence. However to identify these factors it isnecessary to determine the underlying cause of epi-thelial senescence in the prostate and this will requiredirect analysis of senescent epithelial cells in vivo forrelevant biochemical alterations.

It is noteworthy that senescent epithelial cells arepresent in the peripheral zone of the prostate as well asin the transition zone (unpublished observation).However, the mean SA-b gal activity per mg PSA inBPH tissue is 7.78� 3.77 (SEM, n¼ 34) U per mg PSA ascompared to 1.325� 0.37 (SEM, n¼ 30) U per mg PSAin the peripheral zone. This difference is statistically

significant (P¼ 0.036, t-test). It is not clear whether thisalmost 6-fold increase in SA-b gal activity can accountfor the fact that BPH occurs in the transition zone andnot the peripheral zone or whether there are funda-mental differences in the biology of the transition zonethat allow a net increase in proliferation in response togrowth factors and cytokines induced by epithelialsenescence. In addition, other factors such as altera-tions of steroid hormone levels with age, hypoxia, andinflammation [26,27] may also play a role in thepathogenesis of BPH. In summary, while it is clear thatepithelial senescence plays an important role in thepathogenesis of BPH, many questions still remainabout the molecular mechanisms underlying this ex-tremely common disease of older men.

ACKNOWLEDGMENTS

This study is supported by use of the facilities ofthe Houston Department of Veterans Affairs MedicalCenter. P. Castro is supported by a training grant fromNational Institute of Health. The authors thank JosieBeck for her excellent technical assistance.

REFERENCES

1. Campisi J. Replicative senescence: An old lives’ tale? Cell 1996;84:497–500.

2. Campisi J. From cells to organisms: Can we learn about agingfrom cells in culture? Exper Gerontol 2000;36:607–618.

3. Chen QM, Bartholomew JC, Campisi J, Acosta M, Reagan JD,Ames BN. Molecular analysis of H2O2-induced senescent-like growth arrest in normal human fibroblasts: p53 and Rbcontrol G1 arrest but not cell replication. Biochem J 1998;332:43–50.

4. ChenQ, FischerA,Reagan JD,YanLJ,AmesBN.OxidativeDNAdamage and senescence of human diploid fibroblast cells. ProcNatl Acad Sci USA 1995;92:4337–4341.

5. Bringold F, Serrano M. Tumor suppressors and oncogenes incellular senescence. Gerontol 2000;35:317–329.

6. Choi J, Shendrik I, PeacockeM, Peehl D, Buttyan R, Ikeguchi EF,Katz AE, BensonMC. Expression of senescence-associated beta-galactosidase in enlarged prostates from men with benignprostatic hyperplasia. Urology 2000;56:160–166.

7. Castro P, Giri D, Lamb D, IttmannM. Cellular senescence in thepathogenesis of benign prostatic hyperplasia. Prostate 2003;55:30–38.

8. Rinehart CA, Watson JM, Torti VR, Palmieri D. The role ofinterleukin-1 in interactive senescence and age-related humanendometrial cancer. Exp Cell Res 1999;248:599–607.

9. Kumar S, Millis AJ, Baglioni C. Expression of interleukin 1-inducible genes andproduction of interleukin 1 by aging humanfibroblasts. Proc Natl Acad Sci USA 1992;89:4683–4687.

10. Pignolo RJ, Cristofalo VJ, Rotenberg MO. Senescent WI-38 cellsfail to express EPC-1, a gene induced in young cells upon entryinto the G0 state. J Biol Chem 1993;268:8949–8957.

11. Untergasser G, Koch H, Menssen A, Hermeking H. Character-ization of epithelial senescence by serial analysis of gene expres-sion: Identification of genes potentially involved in prostatecancer. Cancer Res 2002;62:6255–6262.

158 Castro et al.

12. Schwarz S, DePrimo S, Grabert L, Fu V, Brooks J, Jerrard D.Novel pathways associated with bypassing cellular senescencein human prostate epithelial cells. J Biol Chem 2002;277:14877–14883.

13. Glynn R, Campion E, Bouchard G, Silbert J. The development ofbenign prostatic hyperplasia among volunteers in the Norma-tive Aging Study. Am J Epidemiol 1985;121:78–83.

14. Giri D, Ittmann M. IL-1a is a paracrine inducer of FGF-7, a keyepithelial growth factor in benign prostatic hyperplasia. Am JPathol 2000;157:249–255.

15. Ropiquet F, Giri D, Lamb D, Ittmann M. FGF7 and FGF2 areincreased in benign prostatic hyperplasia and are associatedwith increased proliferation. J Urol 1999;162:595–599.

16. Giri D, IttmannM. IL-8 is a paracrine inducer of FGF2, a stromaland epithelial growth factor in benignprostatic hyperplasia. AmJ Path 2001;159:139–147.

17. Wheeler T, Lebovitz RM. Fresh tissue harvest for research fromprostatectomy specimens. Prostate 1994;25:274–279.

18. Dimri GP, Lee X, Basile G, Acosta M, Scott G, Roskelley C,MedranoEE, LinskensM,Rubelj I, Pereira-SmithO, PeacockeM,Campisi J. A biomarker that identifies senescent human cells inculture and in aging skin in vivo. Proc Natl Acad Sci USA1995;92:9363–9367.

19. Kim SJ, Uehara H, Karashima T, Mccarty M, Shih N, FidlerIJ. Expression of interleukin-8 correlates with angiogenesis,tumorigenicity, and metastasis of human prostate cancer cellsimplanted orthotopically in nude mice. Neoplasia 2001;3:33–42.

20. Wang E. Senescent human fibroblasts resist programmed celldeath, and failure to suppress bcl2 is involved. Cancer Res 1995;55:2284–2292.

21. Asadi F, Faraj M, Malakouti S, Kukreja SC. Effect of parathyroidhormone related protein, and dihydrotestosterone on prolifera-tion and ornithine decarboxylase mRNA in human prostatecancer cell lines. Int Urol Nephrol 2001;33:417–422.

22. Asadi F, Farraj M, Sharifi R, Malakouti S, Antar S, Kukreja S.Enhanced expression of parathyroid hormone-related protein inprostate cancer as compared with benign prostatic hyperplasia.Hum Pathol 1996;27:1319–1323.

23. Li SL, Goko H, Xu ZD, Kimura G, Sun Y, Kawachi MH, WilsonTG,Wilczynski S, Fujita-YamaguchiY.Expressionof insulin-likegrowth factor (IGF)-II in human prostate, breast, bladder, andparaganglioma tumors. Cell Tissue Res 1998;291:469–479.

24. Monti S, Di Silverio F, Lanzara S, Varasano P, Martini C,Tosti-Croce C, Sciarra F. Insulin-like growth factor-I and -II inhuman benign prostatic hyperplasia: Relationship with bindingproteins 2 and 3 and androgens. Steroids 1998;63:362–366.

25. Slater M, Barden JA, Murphy CR. Changes in growth factorexpression in the ageing prostate may disrupt epithelial–stromal homeostasis. Histochem J 2000;32:357–364.

26. Eaton CL. Aetiology and pathogenesis of benign prostatichyperplasia. Curr Opin Urol 2003;13:7–10.

27. Berger AP, Kofler K, Bektic J, Rogatsch H, Steiner H, Bartsch G,Klocker H. Increased growth factor production in a humanprostatic stromal cell culturemodel caused by hypoxia. Prostate2003;57:57–65.

Interleukin-8 in the Pathogenesis of BPH 159